Re: Help to repair my HP654A Test Oscillator
We all have too many projects! I have a photo to post on the 200CD distortion subject but couldn’t find anything on the web site about whether images are allowed.?
On Sun, Dec 15, 2019 at 1:51 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Jeremy, I couldn't find anything on 200CD in EL34 website.? But the other one yielded essentially the same info as you gave on level setting.
I am resisting the temptation to get into these oscillators.? I don't have bench space and too many things going on.? But soon.
Bob
On Sunday, December 15, 2019, 01:34:00 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
Matching tubes: I didn’t want to repeat myself. For my 200CD, I found a matched pair of 8425A’s (the industrial version of 6AU6A) for V1 and V3. (Please don’t tell HP they’re Tektronix tubes!) For V2 and V4 (6CW5) I found a nice pair of Made-in-Holland Amperex tubes in my spares box.?
I agree the prices on eBay are steep and that seems to be so for much older HP equipment. Clearly, sellers don’t understand that the market is limited for this stuff. The “sold” listings on eBay range from $15.50 to $150. Plus shipping, of course. Flea markets and estate sale prices are much more attractive.?
There is some good 200CD information in <> search for “HP 200CD oscillator repair.”?
Another good link is <>. Under “General Vintage Technology/Success Stories” search for “Hewlett Packard 200CD Wide Range Oscillator.”
Jeremy? N6WFO
On Sun, Dec 15, 2019 at 10:35 AM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I will have to check my serial numbers.? Your most recent message didn't mention matching tubes.
Anyway, this is interesting.? The easiest mods would be to lower the amplitude and improve the lamp sockets.? I will do those first and see what I get.? If results are promising I can get rid of the function generator.? Matching tubes can only work if both units use the same type, or I can find a source of them.? Buying tubes in hopes of finding a good match and then seeing if that helps seems a waste of effort and money.
I have two of these, one is rack mount and I don't want to preserve that one.? So I will concentrate on improving the cabinet mounted unit.
Perusing ebay shows that you can't buy a working unit for less than about $100.? I am amazed that their perceived value is that high.? Once this is done I may put one or both on ebay depending on distortion.? Of course I can brute force lower distortion with a filter but that makes for another project.
Bob K6DDX
On Sunday, December 15, 2019, 10:02:20 AM PST, Jeremy Nichols < jn6wfo@...> wrote:
*** Distortion of the HP-200CD ***
I used my HP-330B Distortion Analyzer to check the 200CD. The 330B is a vacuum tube unit; 0.04% is definitely pushing its limits; it probably would be more fair to claim my 200CD's distortion is "less than 0.1%."
*** Additional information on the HP-200CD Audio Oscillator. ***
1. The later versions using miniature tubes have lower distortion than the earlier versions with octal tubes. The split appears to have taken place at serial prefix 103, which also introduced Dynamic Balance Adjust pot R50 and Hum Balance pot R51.
2. The 200CD was listed in HP catalogs beginning in late 1951 ($150) and was last in the 1985 catalog (un-priced but I have been told it sold for $1250). This 34-year production run is probably a record for HP. The 200CD was HP's second-to-last vacuum tube instrument. One of this nice things about this long run is that there are a zillion of them still floating around for those of us who like to play with the descendants of HP's very first product.
3. The internal pot that sets the maximum output is R11 (1000 Ohms), which is normally set for a 24 Volt no-load output at 1000 Hertz (12 Volts into a 600 Ohm load). If the output is reduced below 10 Volts (loaded) the distortion is reduced; below 7.5 Volts is even better. Clearly, if your application requires higher voltage, you have to live with more distortion.
4. The manual sold by Artek (CD or download) includes in the second volume a group of HP "Service Notes" that are especially valuable for those owning early models using octal tubes. 5. The last revision appears to carry serials prefixed "960-" or "0960A." These final versions use solid-state rectifier diodes and have a standard IEC plug-in power cord.
It's interesting that the 200CD does not have a regulated power supply, only a pi-network filter.
The solid-state rectifier does not make much, if any improvement in distortion. (My "best" 200CD carries serial prefix "333-.")
6. Some excellent information, copied from the old Ampex mailing list, is posted in . This includes detailed information on the 7.5 Volt modification (which I did not do to my 200CD). 7. In the last versions, the sockets for resistor lamps RT1 and RT2 are mounted on a small pc board (with other components). Ensuring good contact between the sockets and the traces can be challenging due to the riveted construction.
Jeremy
On Sat, Dec 14, 2019 at 11:38 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Okay Jeremy I appreciate it.? I have two 200CDs and am sure I can glean a matched pair of tubes from the four.? In any case I am interested in taking this far enough to at least approach the amount of distortion you have.? How did you measure it, and under what settings and conditions?
I have a 331A but don't use it.? A VNA I have does go down into the audio, as well as the FFT in one of my scopes and I can measure distortion that way.? I never measured anything as low as 0.04% though.
Bob K6DDX
On Saturday, December 14, 2019, 11:29:55 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
There are a number of tricks that have been published over the years. I’ll send you copies tomorrow. In the?meantime,? ? balance the two pairs of tubes that make up the push-pull oscillator. The pairs of tubes don’t have to be new/perfect but should be as close as possible in terms of gain.? ? make sure the two lamps in the oscillator feedback have clean bases, are making tight contact with clean sockets, and the sockets have good solder connections. Sockets that are mounted with rivets to a pc board are especially susceptible to poor contact.? ? use the internal gain control (tomorrow I can give you the R-number) to reduce the maximum output to 10 Volts, perhaps even less depending on your needs. Reducing the output voltage lowers distortion.?
I’ll add more in the morning when I’m more awake.?
Jeremy?
On Sat, Dec 14, 2019 at 10:31 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
What's your secret Jeremy?? I have a couple of 200CDs but they are nothing to write home about regarding distortion.
Bob K6DX
On Saturday, December 14, 2019, 10:19:52 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
My650A did fairly well distortion-wise but my favorite is the 200CD. With a few tricks and a little tweaking I was able to get it down to 0.04%, well below spec.
Jeremy
On Sat, Dec 14, 2019 at 9:43 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I know this doesn't help but I did want to say that, in all my years of working with HP gear, this circuit is about the worst design.? I have repaired my 651A a couple of times and have never been able to get it to meet its distortion spec.? Mine works okay but I got tired of messing with it and it's no longer in line for use.? My main repair seemed to center about that FET in the middle of the diagram.? I tried a few different types without much success.? I ddi finally get it to oscillate on all ranges.? The distortion runs a bit over 0.5% worst case and the frequency calibration isn't as good as it could be, but it remains a useful instrument.
Perhaps you'd like to play with it.? The main thing I like is the nice attenuator.? Back in the day, I had a 650A that was a marvelous piece of gear.? Too bad this one didn't match that.
Bob
On Saturday, December 14, 2019, 08:35:49 PM PST, Don Bitters via Groups.Io <donbitters= [email protected]> wrote:
Almost certainly 50Hz AM modulation - 20msec per cycle.? Does touching the 645 chassis change the amplitude of the modulation, if so check the chassis ground at the power cord input, you may have an open ground pin.? If not check the power supplies for 50/100 Hz modulation, percentage of voltage of the supply, the worst looking supply is probably the culprit. Don Bitters
-- Jeremy Nichols
6.
-- Jeremy Nichols
6.
|
Re: OT? Comb mixer, drive level?
Is this for the input receiver section, or the output source section? It seems you're describing an input signal downconverter/sampler. The comb is likely made from the VHF drive power by an SRD, so +20 to 30 dBm is in the right ballpark for level. The output from the SRD then drives the actual input signal mixer to convert down to the IF. The SRD's conversion loss, and the fact that the output power is spread among many harmonics, should still leave enough mixer LO power - perhaps up to a few dBm - at any particular desired harmonic, once it's optimized for it. The instrument's brain should know the correct drive frequency to apply, the harmonic, and the desired IF. It may also control the level into the PA, and the SRD's DC bias, depending on the frequency, so it can get rather complicated. If you have adjusted a tuning network at the PA output, you may have upset things - that may be the matching network for the SRD itself, if it's not built into the comb mixer module. Each SRD is fairly unique in response, so the matching network is adjusted to maximize the efficiency and flatness of the resulting comb output, not necessarily the PA's output. Once you figure out what's supposed to be going on in there (hopefully finding the manual), you may want to step back and see what the adjustments are for.
Ed
|
Re: OT? Comb mixer, drive level?
Have a look at the HP microvawe accessories catalogs around 1970 -1980 : a chapter is devoted to their comb generators. On Sunday, December 15, 2019, 7:07:20 PM GMT+1, Staffan <testjarfalla63@...> wrote:
Hello,
Thanks. Any idea where I can find that manual? Checked ko4bb and Baman, but didn't seem to be there.?
Any idea what power level was used?
Regards,
? Staffan
|
Re: Help to repair my HP654A Test Oscillator
Jeremy, I couldn't find anything on 200CD in EL34 website.? But the other one yielded essentially the same info as you gave on level setting.
I am resisting the temptation to get into these oscillators.? I don't have bench space and too many things going on.? But soon.
Bob
On Sunday, December 15, 2019, 01:34:00 PM PST, Jeremy Nichols <jn6wfo@...> wrote:
Matching tubes: I didn’t want to repeat myself. For my 200CD, I found a matched pair of 8425A’s (the industrial version of 6AU6A) for V1 and V3. (Please don’t tell HP they’re Tektronix tubes!) For V2 and V4 (6CW5) I found a nice pair of Made-in-Holland Amperex tubes in my spares box.?
I agree the prices on eBay are steep and that seems to be so for much older HP equipment. Clearly, sellers don’t understand that the market is limited for this stuff. The “sold” listings on eBay range from $15.50 to $150. Plus shipping, of course. Flea markets and estate sale prices are much more attractive.?
There is some good 200CD information in <> search for “HP 200CD oscillator repair.”?
Another good link is <>. Under “General Vintage Technology/Success Stories” search for “Hewlett Packard 200CD Wide Range Oscillator.”
Jeremy? N6WFO
On Sun, Dec 15, 2019 at 10:35 AM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I will have to check my serial numbers.? Your most recent message didn't mention matching tubes.
Anyway, this is interesting.? The easiest mods would be to lower the amplitude and improve the lamp sockets.? I will do those first and see what I get.? If results are promising I can get rid of the function generator.? Matching tubes can only work if both units use the same type, or I can find a source of them.? Buying tubes in hopes of finding a good match and then seeing if that helps seems a waste of effort and money.
I have two of these, one is rack mount and I don't want to preserve that one.? So I will concentrate on improving the cabinet mounted unit.
Perusing ebay shows that you can't buy a working unit for less than about $100.? I am amazed that their perceived value is that high.? Once this is done I may put one or both on ebay depending on distortion.? Of course I can brute force lower distortion with a filter but that makes for another project.
Bob K6DDX
On Sunday, December 15, 2019, 10:02:20 AM PST, Jeremy Nichols < jn6wfo@...> wrote:
*** Distortion of the HP-200CD ***
I used my HP-330B Distortion Analyzer to check the 200CD. The 330B is a vacuum tube unit; 0.04% is definitely pushing its limits; it probably would be more fair to claim my 200CD's distortion is "less than 0.1%."
*** Additional information on the HP-200CD Audio Oscillator. ***
1. The later versions using miniature tubes have lower distortion than the earlier versions with octal tubes. The split appears to have taken place at serial prefix 103, which also introduced Dynamic Balance Adjust pot R50 and Hum Balance pot R51.
2. The 200CD was listed in HP catalogs beginning in late 1951 ($150) and was last in the 1985 catalog (un-priced but I have been told it sold for $1250). This 34-year production run is probably a record for HP. The 200CD was HP's second-to-last vacuum tube instrument. One of this nice things about this long run is that there are a zillion of them still floating around for those of us who like to play with the descendants of HP's very first product.
3. The internal pot that sets the maximum output is R11 (1000 Ohms), which is normally set for a 24 Volt no-load output at 1000 Hertz (12 Volts into a 600 Ohm load). If the output is reduced below 10 Volts (loaded) the distortion is reduced; below 7.5 Volts is even better. Clearly, if your application requires higher voltage, you have to live with more distortion.
4. The manual sold by Artek (CD or download) includes in the second volume a group of HP "Service Notes" that are especially valuable for those owning early models using octal tubes. 5. The last revision appears to carry serials prefixed "960-" or "0960A." These final versions use solid-state rectifier diodes and have a standard IEC plug-in power cord.
It's interesting that the 200CD does not have a regulated power supply, only a pi-network filter.
The solid-state rectifier does not make much, if any improvement in distortion. (My "best" 200CD carries serial prefix "333-.")
6. Some excellent information, copied from the old Ampex mailing list, is posted in . This includes detailed information on the 7.5 Volt modification (which I did not do to my 200CD). 7. In the last versions, the sockets for resistor lamps RT1 and RT2 are mounted on a small pc board (with other components). Ensuring good contact between the sockets and the traces can be challenging due to the riveted construction.
Jeremy
On Sat, Dec 14, 2019 at 11:38 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Okay Jeremy I appreciate it.? I have two 200CDs and am sure I can glean a matched pair of tubes from the four.? In any case I am interested in taking this far enough to at least approach the amount of distortion you have.? How did you measure it, and under what settings and conditions?
I have a 331A but don't use it.? A VNA I have does go down into the audio, as well as the FFT in one of my scopes and I can measure distortion that way.? I never measured anything as low as 0.04% though.
Bob K6DDX
On Saturday, December 14, 2019, 11:29:55 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
There are a number of tricks that have been published over the years. I’ll send you copies tomorrow. In the?meantime,? ? balance the two pairs of tubes that make up the push-pull oscillator. The pairs of tubes don’t have to be new/perfect but should be as close as possible in terms of gain.? ? make sure the two lamps in the oscillator feedback have clean bases, are making tight contact with clean sockets, and the sockets have good solder connections. Sockets that are mounted with rivets to a pc board are especially susceptible to poor contact.? ? use the internal gain control (tomorrow I can give you the R-number) to reduce the maximum output to 10 Volts, perhaps even less depending on your needs. Reducing the output voltage lowers distortion.?
I’ll add more in the morning when I’m more awake.?
Jeremy?
On Sat, Dec 14, 2019 at 10:31 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
What's your secret Jeremy?? I have a couple of 200CDs but they are nothing to write home about regarding distortion.
Bob K6DX
On Saturday, December 14, 2019, 10:19:52 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
My650A did fairly well distortion-wise but my favorite is the 200CD. With a few tricks and a little tweaking I was able to get it down to 0.04%, well below spec.
Jeremy
On Sat, Dec 14, 2019 at 9:43 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I know this doesn't help but I did want to say that, in all my years of working with HP gear, this circuit is about the worst design.? I have repaired my 651A a couple of times and have never been able to get it to meet its distortion spec.? Mine works okay but I got tired of messing with it and it's no longer in line for use.? My main repair seemed to center about that FET in the middle of the diagram.? I tried a few different types without much success.? I ddi finally get it to oscillate on all ranges.? The distortion runs a bit over 0.5% worst case and the frequency calibration isn't as good as it could be, but it remains a useful instrument.
Perhaps you'd like to play with it.? The main thing I like is the nice attenuator.? Back in the day, I had a 650A that was a marvelous piece of gear.? Too bad this one didn't match that.
Bob
On Saturday, December 14, 2019, 08:35:49 PM PST, Don Bitters via Groups.Io <donbitters= [email protected]> wrote:
Almost certainly 50Hz AM modulation - 20msec per cycle.? Does touching the 645 chassis change the amplitude of the modulation, if so check the chassis ground at the power cord input, you may have an open ground pin.? If not check the power supplies for 50/100 Hz modulation, percentage of voltage of the supply, the worst looking supply is probably the culprit. Don Bitters
-- Jeremy Nichols
6.
|
Re: Help to repair my HP654A Test Oscillator
Matching tubes: I didn’t want to repeat myself. For my 200CD, I found a matched pair of 8425A’s (the industrial version of 6AU6A) for V1 and V3. (Please don’t tell HP they’re Tektronix tubes!) For V2 and V4 (6CW5) I found a nice pair of Made-in-Holland Amperex tubes in my spares box.?
I agree the prices on eBay are steep and that seems to be so for much older HP equipment. Clearly, sellers don’t understand that the market is limited for this stuff. The “sold” listings on eBay range from $15.50 to $150. Plus shipping, of course. Flea markets and estate sale prices are much more attractive.?
There is some good 200CD information in <> search for “HP 200CD oscillator repair.”?
Another good link is <>. Under “General Vintage Technology/Success Stories” search for “Hewlett Packard 200CD Wide Range Oscillator.”
Jeremy? N6WFO
On Sun, Dec 15, 2019 at 10:35 AM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I will have to check my serial numbers.? Your most recent message didn't mention matching tubes.
Anyway, this is interesting.? The easiest mods would be to lower the amplitude and improve the lamp sockets.? I will do those first and see what I get.? If results are promising I can get rid of the function generator.? Matching tubes can only work if both units use the same type, or I can find a source of them.? Buying tubes in hopes of finding a good match and then seeing if that helps seems a waste of effort and money.
I have two of these, one is rack mount and I don't want to preserve that one.? So I will concentrate on improving the cabinet mounted unit.
Perusing ebay shows that you can't buy a working unit for less than about $100.? I am amazed that their perceived value is that high.? Once this is done I may put one or both on ebay depending on distortion.? Of course I can brute force lower distortion with a filter but that makes for another project.
Bob K6DDX
On Sunday, December 15, 2019, 10:02:20 AM PST, Jeremy Nichols < jn6wfo@...> wrote:
*** Distortion of the HP-200CD ***
I used my HP-330B Distortion Analyzer to check the 200CD. The 330B is a vacuum tube unit; 0.04% is definitely pushing its limits; it probably would be more fair to claim my 200CD's distortion is "less than 0.1%."
*** Additional information on the HP-200CD Audio Oscillator. ***
1. The later versions using miniature tubes have lower distortion than the earlier versions with octal tubes. The split appears to have taken place at serial prefix 103, which also introduced Dynamic Balance Adjust pot R50 and Hum Balance pot R51.
2. The 200CD was listed in HP catalogs beginning in late 1951 ($150) and was last in the 1985 catalog (un-priced but I have been told it sold for $1250). This 34-year production run is probably a record for HP. The 200CD was HP's second-to-last vacuum tube instrument. One of this nice things about this long run is that there are a zillion of them still floating around for those of us who like to play with the descendants of HP's very first product.
3. The internal pot that sets the maximum output is R11 (1000 Ohms), which is normally set for a 24 Volt no-load output at 1000 Hertz (12 Volts into a 600 Ohm load). If the output is reduced below 10 Volts (loaded) the distortion is reduced; below 7.5 Volts is even better. Clearly, if your application requires higher voltage, you have to live with more distortion.
4. The manual sold by Artek (CD or download) includes in the second volume a group of HP "Service Notes" that are especially valuable for those owning early models using octal tubes. 5. The last revision appears to carry serials prefixed "960-" or "0960A." These final versions use solid-state rectifier diodes and have a standard IEC plug-in power cord.
It's interesting that the 200CD does not have a regulated power supply, only a pi-network filter.
The solid-state rectifier does not make much, if any improvement in distortion. (My "best" 200CD carries serial prefix "333-.")
6. Some excellent information, copied from the old Ampex mailing list, is posted in . This includes detailed information on the 7.5 Volt modification (which I did not do to my 200CD). 7. In the last versions, the sockets for resistor lamps RT1 and RT2 are mounted on a small pc board (with other components). Ensuring good contact between the sockets and the traces can be challenging due to the riveted construction.
Jeremy
On Sat, Dec 14, 2019 at 11:38 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Okay Jeremy I appreciate it.? I have two 200CDs and am sure I can glean a matched pair of tubes from the four.? In any case I am interested in taking this far enough to at least approach the amount of distortion you have.? How did you measure it, and under what settings and conditions?
I have a 331A but don't use it.? A VNA I have does go down into the audio, as well as the FFT in one of my scopes and I can measure distortion that way.? I never measured anything as low as 0.04% though.
Bob K6DDX
On Saturday, December 14, 2019, 11:29:55 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
There are a number of tricks that have been published over the years. I’ll send you copies tomorrow. In the?meantime,? ? balance the two pairs of tubes that make up the push-pull oscillator. The pairs of tubes don’t have to be new/perfect but should be as close as possible in terms of gain.? ? make sure the two lamps in the oscillator feedback have clean bases, are making tight contact with clean sockets, and the sockets have good solder connections. Sockets that are mounted with rivets to a pc board are especially susceptible to poor contact.? ? use the internal gain control (tomorrow I can give you the R-number) to reduce the maximum output to 10 Volts, perhaps even less depending on your needs. Reducing the output voltage lowers distortion.?
I’ll add more in the morning when I’m more awake.?
Jeremy?
On Sat, Dec 14, 2019 at 10:31 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
What's your secret Jeremy?? I have a couple of 200CDs but they are nothing to write home about regarding distortion.
Bob K6DX
On Saturday, December 14, 2019, 10:19:52 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
My650A did fairly well distortion-wise but my favorite is the 200CD. With a few tricks and a little tweaking I was able to get it down to 0.04%, well below spec.
Jeremy
On Sat, Dec 14, 2019 at 9:43 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I know this doesn't help but I did want to say that, in all my years of working with HP gear, this circuit is about the worst design.? I have repaired my 651A a couple of times and have never been able to get it to meet its distortion spec.? Mine works okay but I got tired of messing with it and it's no longer in line for use.? My main repair seemed to center about that FET in the middle of the diagram.? I tried a few different types without much success.? I ddi finally get it to oscillate on all ranges.? The distortion runs a bit over 0.5% worst case and the frequency calibration isn't as good as it could be, but it remains a useful instrument.
Perhaps you'd like to play with it.? The main thing I like is the nice attenuator.? Back in the day, I had a 650A that was a marvelous piece of gear.? Too bad this one didn't match that.
Bob
On Saturday, December 14, 2019, 08:35:49 PM PST, Don Bitters via Groups.Io <donbitters= [email protected]> wrote:
Almost certainly 50Hz AM modulation - 20msec per cycle.? Does touching the 645 chassis change the amplitude of the modulation, if so check the chassis ground at the power cord input, you may have an open ground pin.? If not check the power supplies for 50/100 Hz modulation, percentage of voltage of the supply, the worst looking supply is probably the culprit. Don Bitters
-- Jeremy Nichols
6.
|
Re: seeking schematic and any experienced advice FOLLOW-UP MESSAGE with correct model number
Hi,
The Vega 450 datasheet and manual are available online and the supply should run OK without any load. Checking the supply out of the scope is best approach. I do have a spare Vega 450 with 2 low voltage and 2 12V output modules but would have to check the currents.
Robert G8RPI.
|
Re: seeking schematic and any experienced advice FOLLOW-UP MESSAGE with correct model number
From personal experience with that flavour of SMPS in other products.??
It's more likely that the PFC front end has "blown up!"
They are fairly easy to remove the PCB's, upon which you'll easily see
the tell-tale marks of disaster.
I had one though, where one of the regulator cards had gone short-cct at
it's HVDC input, destroying itself, and taking the input card too (as
well as blowing one of the interconnect pins, like a fuse-wire!)
Like many similar things, there is a lot of current going round in
circles in those PSU's, and any compromise in cooling seems to push them
over the edge, even if the load is well within spec'.
TDK Lambda in the UK are reachable, and have a repair service, should
you need.? I don't know if they have similar facilities elsewhere on the
planet, other than perhaps the far east, where a lot of the modules are
actually manufactured.
73.
Dave B.
--
Created on and sent from a Unix like PC running and using free and open source software:
|
Re: 8350B w/83590A repair advice
On Sun, 15 Dec 2019, Paul Bicknell wrote: No problems, everybody else also welcome to download it. Thank you for the manual
Best Regards Paul
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sergey
Kubushyn
Sent: 15 December 2019 03:08
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8350B w/83590A repair advice
On Sun, 15 Dec 2019, Paul Bicknell wrote:
Put it on my web site at:
Not of stellar quality but pretty decent, with schematics, drawings, parts
lists and so on.
Hi do you have the service manual for the Wiltron 6669B
As I have a 6669A
Paul
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Sergey
Kubushyn
Sent: 15 December 2019 01:45
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8350B w/83590A repair advice
On Sat, 14 Dec 2019, peter bunge wrote:
Check which unit has the problems.
Best done with a plug-in you can trust. Don't tinker until you know which
is the bad unit. It is almost certainly the plug-in. 8350B itself is a simple device having
almost no influence on plug-in performance.
Those 8350B plug-ins are very tricky to adjust and calibrate. Their output
is a product of several oscillators and filters that all must be tuned
properly. The most difficult is 83592B (and other wide range devices) LOW
frequency range that is a product of 2 several-GHz oscillators so it is
EXTREMELY sensitive to proper adjustments -- you have one of those very
slightly off and there is no lowest band output at all. Zero. The same is
true for other ranges but sensitivity is much lower so relatively big
misalignment would still produce SOME output.
I spent a week trying to calibrate 83592B and put it aside for a time being
still not fully calibrated. It is very complex device and different settings
all affect each other so you have to re-do calibration multiple times.
I've got two 8340B instruments since that so I lost all interest in 8350B
and its plugins so they are gathering dust right now. I've also got Wiltron
6669B that goes up to 40GHz that works somehow but it is the same kind of
instrument as 8350B, just with everything from all possible plug-ins built
into single instrument so it also begs for proper alignment/calibration. It
works but produces something far from sinewave on the lowest band so it
definitely has to be re-adjusted from scratch.
BTW, if somebody is interested I can list those 2 fully working and
calibrated 8350B mainframes I have as well as fully working fully
refurbished and calibrated 83570A plug-in as well as several different
plug-ins in various refurbisment/adjustment/calibration states and some
partial units for parts as I don't need them anymore and have no time to
waste on them. I do also have printed manuals on almost all of those that I
would also rather get rid of. Haven't done it so far as there are too much
work on my calendar so I'd rather spend time on something else but if
somebody is interested I can make a list and either sell those direct or
list them on EBay.
BTW, there is little sense in putting new batteries in 8350B (although I
did) as they back up RAM that only holds last settings, nothing else. There
are no calibration constants there or something else that might've been of
any importance. It will just power-up in a default state if batteries are
dead or missing, nothing else. All settings are pure mechanical adjustments
that don't require any battery to persist. --- * * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * *
|
Re: Help to repair my HP654A Test Oscillator
I will have to check my serial numbers.? Your most recent message didn't mention matching tubes.
Anyway, this is interesting.? The easiest mods would be to lower the amplitude and improve the lamp sockets.? I will do those first and see what I get.? If results are promising I can get rid of the function generator.? Matching tubes can only work if both units use the same type, or I can find a source of them.? Buying tubes in hopes of finding a good match and then seeing if that helps seems a waste of effort and money.
I have two of these, one is rack mount and I don't want to preserve that one.? So I will concentrate on improving the cabinet mounted unit.
Perusing ebay shows that you can't buy a working unit for less than about $100.? I am amazed that their perceived value is that high.? Once this is done I may put one or both on ebay depending on distortion.? Of course I can brute force lower distortion with a filter but that makes for another project.
Bob K6DDX
On Sunday, December 15, 2019, 10:02:20 AM PST, Jeremy Nichols <jn6wfo@...> wrote:
*** Distortion of the HP-200CD ***
I used my HP-330B Distortion Analyzer to check the 200CD. The 330B is a vacuum tube unit; 0.04% is definitely pushing its limits; it probably would be more fair to claim my 200CD's distortion is "less than 0.1%."
*** Additional information on the HP-200CD Audio Oscillator. ***
1. The later versions using miniature tubes have lower distortion than the earlier versions with octal tubes. The split appears to have taken place at serial prefix 103, which also introduced Dynamic Balance Adjust pot R50 and Hum Balance pot R51.
2. The 200CD was listed in HP catalogs beginning in late 1951 ($150) and was last in the 1985 catalog (un-priced but I have been told it sold for $1250). This 34-year production run is probably a record for HP. The 200CD was HP's second-to-last vacuum tube instrument. One of this nice things about this long run is that there are a zillion of them still floating around for those of us who like to play with the descendants of HP's very first product.
3. The internal pot that sets the maximum output is R11 (1000 Ohms), which is normally set for a 24 Volt no-load output at 1000 Hertz (12 Volts into a 600 Ohm load). If the output is reduced below 10 Volts (loaded) the distortion is reduced; below 7.5 Volts is even better. Clearly, if your application requires higher voltage, you have to live with more distortion.
4. The manual sold by Artek (CD or download) includes in the second volume a group of HP "Service Notes" that are especially valuable for those owning early models using octal tubes. 5. The last revision appears to carry serials prefixed "960-" or "0960A." These final versions use solid-state rectifier diodes and have a standard IEC plug-in power cord.
It's interesting that the 200CD does not have a regulated power supply, only a pi-network filter.
The solid-state rectifier does not make much, if any improvement in distortion. (My "best" 200CD carries serial prefix "333-.")
6. Some excellent information, copied from the old Ampex mailing list, is posted in . This includes detailed information on the 7.5 Volt modification (which I did not do to my 200CD). 7. In the last versions, the sockets for resistor lamps RT1 and RT2 are mounted on a small pc board (with other components). Ensuring good contact between the sockets and the traces can be challenging due to the riveted construction.
Jeremy
On Sat, Dec 14, 2019 at 11:38 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Okay Jeremy I appreciate it.? I have two 200CDs and am sure I can glean a matched pair of tubes from the four.? In any case I am interested in taking this far enough to at least approach the amount of distortion you have.? How did you measure it, and under what settings and conditions?
I have a 331A but don't use it.? A VNA I have does go down into the audio, as well as the FFT in one of my scopes and I can measure distortion that way.? I never measured anything as low as 0.04% though.
Bob K6DDX
On Saturday, December 14, 2019, 11:29:55 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
There are a number of tricks that have been published over the years. I’ll send you copies tomorrow. In the?meantime,? ? balance the two pairs of tubes that make up the push-pull oscillator. The pairs of tubes don’t have to be new/perfect but should be as close as possible in terms of gain.? ? make sure the two lamps in the oscillator feedback have clean bases, are making tight contact with clean sockets, and the sockets have good solder connections. Sockets that are mounted with rivets to a pc board are especially susceptible to poor contact.? ? use the internal gain control (tomorrow I can give you the R-number) to reduce the maximum output to 10 Volts, perhaps even less depending on your needs. Reducing the output voltage lowers distortion.?
I’ll add more in the morning when I’m more awake.?
Jeremy?
On Sat, Dec 14, 2019 at 10:31 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
What's your secret Jeremy?? I have a couple of 200CDs but they are nothing to write home about regarding distortion.
Bob K6DX
On Saturday, December 14, 2019, 10:19:52 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
My650A did fairly well distortion-wise but my favorite is the 200CD. With a few tricks and a little tweaking I was able to get it down to 0.04%, well below spec.
Jeremy
On Sat, Dec 14, 2019 at 9:43 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I know this doesn't help but I did want to say that, in all my years of working with HP gear, this circuit is about the worst design.? I have repaired my 651A a couple of times and have never been able to get it to meet its distortion spec.? Mine works okay but I got tired of messing with it and it's no longer in line for use.? My main repair seemed to center about that FET in the middle of the diagram.? I tried a few different types without much success.? I ddi finally get it to oscillate on all ranges.? The distortion runs a bit over 0.5% worst case and the frequency calibration isn't as good as it could be, but it remains a useful instrument.
Perhaps you'd like to play with it.? The main thing I like is the nice attenuator.? Back in the day, I had a 650A that was a marvelous piece of gear.? Too bad this one didn't match that.
Bob
On Saturday, December 14, 2019, 08:35:49 PM PST, Don Bitters via Groups.Io <donbitters= [email protected]> wrote:
Almost certainly 50Hz AM modulation - 20msec per cycle.? Does touching the 645 chassis change the amplitude of the modulation, if so check the chassis ground at the power cord input, you may have an open ground pin.? If not check the power supplies for 50/100 Hz modulation, percentage of voltage of the supply, the worst looking supply is probably the culprit. Don Bitters
-- Jeremy Nichols
6.
|
Argh, doesn't appear to be a regular parity, sum complement, or CRC. Need to pull the EEPROM from the other 54712A (with a different serial number) and XOR with this one, to see where the differences appear. Hopefully, there is a checksum over the header only, up to 0x80 (that is, the calibration data in not included in the checksum). That's easy to check by modifying one bit in the calibration data area and watching if the module is still recognized (or not). The big concern is the number of solders on the 54712A board, don't want to kill the pads. I'm using hot air, but still a risk.
|
Re: OT? Comb mixer, drive level?
Hello,
Thanks. Any idea where I can find that manual? Checked ko4bb and Baman, but didn't seem to be there.?
Any idea what power level was used?
Regards,
? Staffan
|
Re: Help to repair my HP654A Test Oscillator
*** Distortion of the HP-200CD ***
I used my HP-330B Distortion Analyzer to check the 200CD. The 330B is a vacuum tube unit; 0.04% is definitely pushing its limits; it probably would be more fair to claim my 200CD's distortion is "less than 0.1%."
*** Additional information on the HP-200CD Audio Oscillator. ***
1. The later versions using miniature tubes have lower distortion than the earlier versions with octal tubes. The split appears to have taken place at serial prefix 103, which also introduced Dynamic Balance Adjust pot R50 and Hum Balance pot R51.
2. The 200CD was listed in HP catalogs beginning in late 1951 ($150) and was last in the 1985 catalog (un-priced but I have been told it sold for $1250). This 34-year production run is probably a record for HP. The 200CD was HP's second-to-last vacuum tube instrument. One of this nice things about this long run is that there are a zillion of them still floating around for those of us who like to play with the descendants of HP's very first product.
3. The internal pot that sets the maximum output is R11 (1000 Ohms), which is normally set for a 24 Volt no-load output at 1000 Hertz (12 Volts into a 600 Ohm load). If the output is reduced below 10 Volts (loaded) the distortion is reduced; below 7.5 Volts is even better. Clearly, if your application requires higher voltage, you have to live with more distortion.
4. The manual sold by Artek (CD or download) includes in the second volume a group of HP "Service Notes" that are especially valuable for those owning early models using octal tubes. 5. The last revision appears to carry serials prefixed "960-" or "0960A." These final versions use solid-state rectifier diodes and have a standard IEC plug-in power cord.
It's interesting that the 200CD does not have a regulated power supply, only a pi-network filter.
The solid-state rectifier does not make much, if any improvement in distortion. (My "best" 200CD carries serial prefix "333-.")
6. Some excellent information, copied from the old Ampex mailing list, is posted in . This includes detailed information on the 7.5 Volt modification (which I did not do to my 200CD). 7. In the last versions, the sockets for resistor lamps RT1 and RT2 are mounted on a small pc board (with other components). Ensuring good contact between the sockets and the traces can be challenging due to the riveted construction.
Jeremy
On Sat, Dec 14, 2019 at 11:38 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
Okay Jeremy I appreciate it.? I have two 200CDs and am sure I can glean a matched pair of tubes from the four.? In any case I am interested in taking this far enough to at least approach the amount of distortion you have.? How did you measure it, and under what settings and conditions?
I have a 331A but don't use it.? A VNA I have does go down into the audio, as well as the FFT in one of my scopes and I can measure distortion that way.? I never measured anything as low as 0.04% though.
Bob K6DDX
On Saturday, December 14, 2019, 11:29:55 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
There are a number of tricks that have been published over the years. I’ll send you copies tomorrow. In the?meantime,? ? balance the two pairs of tubes that make up the push-pull oscillator. The pairs of tubes don’t have to be new/perfect but should be as close as possible in terms of gain.? ? make sure the two lamps in the oscillator feedback have clean bases, are making tight contact with clean sockets, and the sockets have good solder connections. Sockets that are mounted with rivets to a pc board are especially susceptible to poor contact.? ? use the internal gain control (tomorrow I can give you the R-number) to reduce the maximum output to 10 Volts, perhaps even less depending on your needs. Reducing the output voltage lowers distortion.?
I’ll add more in the morning when I’m more awake.?
Jeremy?
On Sat, Dec 14, 2019 at 10:31 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
What's your secret Jeremy?? I have a couple of 200CDs but they are nothing to write home about regarding distortion.
Bob K6DX
On Saturday, December 14, 2019, 10:19:52 PM PST, Jeremy Nichols < jn6wfo@...> wrote:
My650A did fairly well distortion-wise but my favorite is the 200CD. With a few tricks and a little tweaking I was able to get it down to 0.04%, well below spec.
Jeremy
On Sat, Dec 14, 2019 at 9:43 PM Bob Albert via Groups.Io <bob91343= [email protected]> wrote:
I know this doesn't help but I did want to say that, in all my years of working with HP gear, this circuit is about the worst design.? I have repaired my 651A a couple of times and have never been able to get it to meet its distortion spec.? Mine works okay but I got tired of messing with it and it's no longer in line for use.? My main repair seemed to center about that FET in the middle of the diagram.? I tried a few different types without much success.? I ddi finally get it to oscillate on all ranges.? The distortion runs a bit over 0.5% worst case and the frequency calibration isn't as good as it could be, but it remains a useful instrument.
Perhaps you'd like to play with it.? The main thing I like is the nice attenuator.? Back in the day, I had a 650A that was a marvelous piece of gear.? Too bad this one didn't match that.
Bob
On Saturday, December 14, 2019, 08:35:49 PM PST, Don Bitters via Groups.Io <donbitters= [email protected]> wrote:
Almost certainly 50Hz AM modulation - 20msec per cycle.? Does touching the 645 chassis change the amplitude of the modulation, if so check the chassis ground at the power cord input, you may have an open ground pin.? If not check the power supplies for 50/100 Hz modulation, percentage of voltage of the supply, the worst looking supply is probably the culprit. Don Bitters
-- Jeremy Nichols
6.
|
Re: 8350B w/83590A repair advice
The early 8970A Noise Figure Meters used the same battery pack for the same purpose though it appeared to serve a larger number of RAM's than the 8350B.
Later, the NF Meters were modified to eliminate the battery pack in favor of a 3 VDC Lithium coin cell by adding a series diode to prevent the 'charging circuit' from 'charging' the Lithium battery.
These changes are outlined in Change C included in the Operating and Service Manual, P/N 08970-90015, for the 8970A. I can send a scan of the relevant pages if anyone is interested.
Has anyone tried this modification with the 8350B's? It would seem that it might work, though in the 8350B manual, it refers to a '5VDC Battery'.
Joe
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of Chuck Harris Sent: Saturday, December 14, 2019 11:17 PM To: [email protected]Subject: Re: [HP-Agilent-Keysight-equipment] 8350B w/83590A repair advice After I clean up and repair the damage to the boards on either side of the pack, I remove the pack "socket" and wires, and throw it all away. There is no significant reason to keep it, as it only saves the front panel settings... which aren't all that complicated. -Chuck Harris Paul Bicknell wrote: Hi Chuck
That reminds me
I must order some new cells as all of my 8350 are running without any
batteries
Ok winter job I have collected a truck load of 8620 10 off
so must service and make good should salvage 8 out of the 10
Regards Paul
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Chuck Harris
Sent: 14 December 2019 15:45
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8350B w/83590A repair advice
Note that the 8350B has a 3- AA cell nicad battery,
repurposed from the HP-35 calculators. The pack is
on trickle charge whenever the 8350B is turned on,
and after a year, or two of that abuse, spews its
highly caustic guts all over the side of whatever
plugin is in the 8350B at the time.
The electrolyte tends to ruin the GPIB board, and
the plugin's main circuit board.
You might look there.
-Chuck Harris
|
Re: OT? Comb mixer, drive level?
Have a look at the Ailtech 757 where he first mixer is an HP comb generator driven beyond its normal specs. On Sunday, December 15, 2019, 5:33:12 PM GMT+1, Staffan <testjarfalla63@...> wrote:
Hello,
I'm working on fixing a VNA with a comb mixer (input 360 - 540 MHz) as the first stage, covering 270 MHz - 20 GHz (sic). It seems the driver amp was off tuning and only output approximately 22-23 dBm at best.
Finding out that it was the driver was the tricky thing, retuning it was maybe too simple. It now outputs 30-31 dBm over most of the range and the question is: can this power level be too high for the mixer input? I see no simple way of controlling output power.?
The schematics of the PA (output stage) is fairly similar to the amp in the comb generator of the 8569A (schematics A42).
The reason for the question to be off topic is that the driver sits in a Wiltron 37247A VNA...
I will ask in the Wiltron group also, but have hope that someone here may be able to tell whether this _type_ of mixers typically handle these levels or if I should (how?) reduce the output level. Any hints are very much appreciated. I - unfortunately - have no HP VNA so I don't know where to look, but guess instruments from the 80s and 90s generally were built this way? Anyone knows an HP unit that may have the same type of mixer and if so, its input drive level?
I'm keen on installing the amp, but would hate to destroy the following mixer just because I was too quick on testing...
In my world one watt is very high power. Googling SRD mixers indicate that these are OK levels, but perhaps in the upper range?
Again, any comments welcome!
Regards,
? Staffan
|
Re: Wanted: HP 3458a DVM in non-working condition
Rick,
Be prepared to pay big $$$ unless you happen to luck out and find a bargin.
Even a non-working 3458A goes for over $1000.
This instrument is rarely found in a non-working condition.
The general failure after 20 years is the CAL RAM battery running out of juice.
Good luck.
|
OT? Comb mixer, drive level?
Hello,
I'm working on fixing a VNA with a comb mixer (input 360 - 540 MHz) as the first stage, covering 270 MHz - 20 GHz (sic). It seems the driver amp was off tuning and only output approximately 22-23 dBm at best.
Finding out that it was the driver was the tricky thing, retuning it was maybe too simple. It now outputs 30-31 dBm over most of the range and the question is: can this power level be too high for the mixer input? I see no simple way of controlling output power.?
The schematics of the PA (output stage) is fairly similar to the amp in the comb generator of the 8569A (schematics A42).
The reason for the question to be off topic is that the driver sits in a Wiltron 37247A VNA...
I will ask in the Wiltron group also, but have hope that someone here may be able to tell whether this _type_ of mixers typically handle these levels or if I should (how?) reduce the output level. Any hints are very much appreciated. I - unfortunately - have no HP VNA so I don't know where to look, but guess instruments from the 80s and 90s generally were built this way? Anyone knows an HP unit that may have the same type of mixer and if so, its input drive level?
I'm keen on installing the amp, but would hate to destroy the following mixer just because I was too quick on testing...
In my world one watt is very high power. Googling SRD mixers indicate that these are OK levels, but perhaps in the upper range?
Again, any comments welcome!
Regards,
? Staffan
|
Re: Does anyone have any PIC17C42A-16I/L to spare (ideally in UK)
Thanks – I tried to order via their website but failed (?).? I’ll ?? them on Monday ? David ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Robert G8RPI via Groups.Io
Sent: 15 December 2019 15:27
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] Does anyone have any PIC17C42A-16I/L to spare (ideally in UK) ?
Microchip Direct have over 1000 PIC17C42A-25I/L in stock at ?7.21 each. The hgher speed should not be an issue.
?https://www.microchipdirect.com/product/search/all/PIC17C42A?facet=on&facet=true&fq={!tag=PT}PackageType_s:(%22PLCC%22)&facet.field=PackageType_s&start=0&rows=50
|
Re: Does anyone have any PIC17C42A-16I/L to spare (ideally in UK)
Microchip Direct have over 1000 PIC17C42A-25I/L in stock at ?7.21 each. The hgher speed should not be an issue.
?https://www.microchipdirect.com/product/search/all/PIC17C42A?facet=on&facet=true&fq={!tag=PT}PackageType_s:(%22PLCC%22)&facet.field=PackageType_s&start=0&rows=50
|
Re: Does anyone have any PIC17C42A-16I/L to spare (ideally in UK)
Microchip never obsoletes a product or so they told me 10 years ago. If that is true you should be able to get one direct. Chuck
|
Re: 8350B w/83590A repair advice
Hi Mark
?
From the results given it could be that as
stated earlier it requires realigning
The master yig is multiplied several times
to give the full band?
?
so during alignment the switching between the
bands requires setting so they do not overlap or leave gaps
and it is this measurement you require to monitor
the sweep with a scalar analyser ?
?
but in my early days I used a diode detector
into a scope with the sweeper supplying the X axes
this is best dun with an attenuate and not
levelled as the bands stand out easier
?
during the alignment you might have to
change resisters? this will require the use of a decade resistor box
sorry its been over 25 years since I did
one
?
Paul
?
Unit just arrived.
Initial testing does not agree with original complaint: "Output at 2Ghz is -10db below what it should be. Won't
produce a signal much past 10Ghz and at 10Ghz it is -30db below the RF output
level setting. No output past this frequency"
Output is approximately 5 dB below stated level, from 2
to 20 GHz. At the AUX output, I measure also 5 dB low, or -5 dBm, rather than 0
dBm.
With the ALC disabled, output is around 18 dBm, which
might also be low (not sure).
So it seems I'm about 5 dB low across the board. Maybe
YIG output is weak OR perhaps the MOD/Coupler...?
At any rate, may not be as bad as expected.
Next, open 'er up!
Mark
No virus found in this message.
Checked by AVG -
Version: 2016.0.8048 / Virus Database: 4793/15886 - Release Date: 08/14/18
Internal Virus Database is out of date.
|
Jeremy Nichols
Sergey Kubushyn
David C. Partridge
Paul Bicknell