And yet the scope is stable enough already.
There is an old saying: "If it ain't broke, don't
fix it."
-Chuck Harris
Jim Olson wrote:
toggle quoted message
Show quoted text
Forgot to add that the higher class 2 have much less change to capacitance +-15% vs +20%-85% so should make the scope more stable.
Jim
|
Forgot to add that the higher class 2 have much less change to capacitance +-15% vs +20%-85% so should make the scope more stable.
Jim
|
Here is a pondering question for you all to think on. How many of you have considered replacing (upgrade) the ceramic disc caps to a higher class 2 grade like X5R or X7R to replace all the Z5U ones used as they are the worst of the class 2 and as they deteriorate with age plus are susceptible to temp and frequency changes to capacitance so they might be responsible for the little niggling problems with the scopes over time or after repairs but it is just not right?
After all they are old and older technology and in most cases much used. Just a thought to consider.
Jim
|
On Tue, Jul 24, 2018 at 01:02 PM, Bert Haskins wrote: Be careful about using a fet with a higher voltage rating, as the voltage
rating goes up, so does the the gate capacitance Yes forgot to tell him to check for that too... but in my defense it was implied in my generic "meets or exceeds specs" general statement ^^ Of course in case of gate capacitance, "exceeds" means lower not higher... just in case ! ^^ Luckily he would probably have been fine anyway, unless going wild on over rating the voltage. Vince
|
On 7/24/2018 12:27 PM, satbeginner wrote: Hi all,
a cheap and broken 2215A found it's way to my home :-), so now I am in the process of getting it repaired.
I used this document from H?kan that helped me a lot:
The fuse would blow instantly, so I separated the main PS from the rest by removing Q9070.
An inspection by the eye learned that there were several low voltage elco's leaking,
C960/C962 (+8,6V) were leaking;
C961/C963 (-8,6V) were leaking;
C968/C970 (+5,2V) were bad;
C956 (+30V) was leaking;
Fortunately no real damage to the PCB.
Also the Q9070 and CR907 were toast.
According to the document the FET is a IRF730 (400V 5,5A 0,75Ohm 12ns) and the diode a BYD73G (400V 1A 50ns)
I replaced all the capacitors first, and now the scope has trace(s) when applying 43VDC to TP940 (Pos) and TP950 (neg).
Being the optimist, I replaced Q9070 with a FQPF4N90C (900V 4A 3,5Ohm 50ns) and CR907 with a PR1507 (1000V 1,5A 300ns) I harvested from a switching PS,
but.... They last about 2 seconds...
Based on H?kan's document I applied a test voltage to C925, and the signals at the chip U930 check out, and since I tested all surrounding components Q908, CR908, R909, CR920, I come to think it must be the speed of the Q9070 and CR907 that causes the sudden death of these replacement components.
My question is: does anybody (recently) repaired a 2215A LVPS and replaced these 2 comonents?
If so, what replacements were used ?
The documents suggest a MTP6N55 for the FET and a BYD73G for the diode, but I would love to hear if there are other successful candidates.
Pictures can be found in this album: /g/TekScopes/album?id=64919
un saludo,
Leo
Be careful about using a fet with a higher voltage rating, as the voltage rating goes up, so does the the gate capacitance and the 22xx drive circuit is marginal to say the least. Check drive circuit carefully! I have learned the hard way to disconnect the voltage multiplier early in the troubleshooting procedure. A bad voltage multiplier puts a large load on the rest of the power supply. Get all the low voltages right first. -Bert
|
Re: Tek 495P A54 Memory Board Repair
Unless it's an expensive IC, troubleshooting a circuit like that is difficult since it is only read once while the SA is booting up. That's why most people would just change a gumdrop IC to attempt the repair. There isn't much else that could cause an error like that.
I used to spend full days testing and repairing MC68340 based embedded controller boards That IC is part of the Motorola 68000 family.
I am in Florida. I am looking at a portable shed to use as a new shop until the 1200 square foot shop is usable.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Rick Boswell <frboswell@...>
Sent: Jul 24, 2018 2:48 PM
To: [email protected]
Subject: Re: [TekScopes] Tek 495P A54 Memory Board Repair
Well ¡, I don't yet know that it needs to be replaced. The buffer IC may not be the problem and I don't want to begin changing parts until I have a more conclusive idea as to which part to change. Located NH. What I am looking for per first post is someone who can trouble shoot & repair, or supply a known good board.
RB
|
Re: Tek 495P A54 Memory Board Repair
To be fair, some of the members are just getting into electronics. I've been at it for over 50 years. At one point I spent eight hours a day testing new circuit boards and repairing the bad ones. My work was better the the two women in rework, and they had better toys than my couple Ungar Loner irons.
For my own work, I prefer a vacuum desoldering iron. In days past I replaced thousands of bad DIP ICs with a Radio Shack desoldering iron. I did use some wick for desoldering, but I used regular copper braid and liquid RMA flux because it does less damage. NASA referred to it as 'wet wicking' and recommended it because it caused less damage to circuit boards. Also, when you trim the used part of the braid, leave about half the width of the wick of used wick. This allows you to transfer the heat faster, and starts the capillary action faster.
I recovered and sold thousands of 256K RAM chips back in the 286 computer days with a solder pot. We sold them for $2.75 each, and we had people begging for them because there was a world wide shortage on the spot market. We would get boards with 144 of them, and I could remove all of them from a board in under five minutes with minimal or no damage to the PCB.I used surface tension to have the solder slightly higher than the edges of the solder pot, and a special pair of pliers or a long, thin stainless steel rod that was ground flat at the end to slip it under the ICs.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Sergey Kubushyn <ksi@...>
Sent: Jul 24, 2018 3:07 PM
To: [email protected]
Subject: Re: [TekScopes] Tek 495P A54 Memory Board Repair
On Tue, 24 Jul 2018, Michael A. Terrell wrote:
It is simply astonishing how helpless some people are :)
If one wants to replace a DIP IC he doesn't need _ANY_ specialized
desoldering equipment. Just cut all pins as close to the case as possible,
throw away that brick and pull the remaining pin stubs one by one using
soldering iron and tweezer. Then use desoldering braid to clean the holes
from remaining solder and the board is as good as new ready for a
replacement IC. No special desoldering tools needed and no such tools make
better job.
The only case when one would require such tools is removing an IC undamaged
for further inspection or reuse. In his case it is totally unnecessary
because replacement buffer IC cost pennies so it is not even worth bothering
with checking a suspicious one -- just yank it out and replace with a new
one.
Where are you located? Someone on the group may live nearby. I would offer
to replace it for you, but I can't get to my workbench until I finish
disposing of the things my family stored in the building.
Michael A. Terrell
|
Hi Vincent,
this is funny, after my previous post I went looking for a Faster, Lower On-resistance, Higher Voltage FET, and found this one: IPA60R280E6 , so this is in the same area as the one you advised and used.. :-) :-)
From the same seller I ordered some ultra fast diodes ER506 (600V 5A 35ns) as replacement for CR907, so these should work too.
BTW, in the beginning my choosen FET and diode did "work", but that was because C956 still had a short, so the neon's were blinking.
It never had to deliver the actual power to the scope.
After I fixed C956, the primary PS died instantly because of the wrong FET.
Thanks for the advise, will keep you posted as soon as the parts arrive.
(I am in inland Spain, so be patient)
Leo
|
Re: Tek 495P A54 Memory Board Repair
On Tue, 24 Jul 2018, Michael A. Terrell wrote: It is simply astonishing how helpless some people are :) If one wants to replace a DIP IC he doesn't need _ANY_ specialized desoldering equipment. Just cut all pins as close to the case as possible, throw away that brick and pull the remaining pin stubs one by one using soldering iron and tweezer. Then use desoldering braid to clean the holes from remaining solder and the board is as good as new ready for a replacement IC. No special desoldering tools needed and no such tools make better job. The only case when one would require such tools is removing an IC undamaged for further inspection or reuse. In his case it is totally unnecessary because replacement buffer IC cost pennies so it is not even worth bothering with checking a suspicious one -- just yank it out and replace with a new one. Where are you located? Someone on the group may live nearby. I would offer
to replace it for you, but I can't get to my workbench until I finish
disposing of the things my family stored in the building.
Michael A. Terrell
-----Original Message-----
From: Rick Boswell <frboswell@...>
Sent: Jul 24, 2018 1:38 PM
To: [email protected]
Subject: Re: [TekScopes] Tek 495P A54 Memory Board Repair
The switches are enabled to the data bus via tri-state buffer and an
enable signal that occurs just once during boot up. Challenging to
trouble shoot with my limited test equipment. And, if the buffer is bad,
I lack the desoldering gear necessary to change it. --- * * KSI@home KOI8 Net < > The impossible we do immediately. * * Las Vegas NV, USA < > Miracles require 24-hour notice. * *
|
Re: Tek 495P A54 Memory Board Repair
Well ¡, I don't yet know that it needs to be replaced. The buffer IC may not be the problem and I don't want to begin changing parts until I have a more conclusive idea as to which part to change. Located NH. What I am looking for per first post is someone who can trouble shoot & repair, or supply a known good board.
RB
|
Hi Leo,
I fixed a couple of 22XX SMSP. The one in my 2232 last summer, and the other day the one in my 2215.
The regulator is not too fussy as to what FET model you use to replace the old one.... AS LONG as it meets or exceed the specs of the old one, of course ! ;-P Obviously the FEt you used did not meet the specs, by a long shot, so no big surprised that it failed.
Typically you just pick whatever FET you can get from your favorite parts supplier, that meets the specs( voltage and current rating), and as for Rdson, typically modern FET have a much, much lower Rdson, in the order of tens of milli-ohms versus a few ohms for the old ones.
For my 2215, the old FET was an ancient IRF820, 3ohms, horrible. I ran a parametric search and ended picking up a FET made " Infineon ", part number : IPP60R280P7, which is 0.028ohms... yes, two orders of magnitude lower. Great because the FET is not mounted on a heat sink, there is zero airflow, not even still air around it (because it's trapped in ins plastic holder), and this particular scope has no cooling fan to top it all. So really, the much lower Rdson is welcome.
Anyway, this FET works just fine..
In my 2232 I used another FET model and brand, don't even remember what it was, but works just fine also. An IRF something maybe...
Your scope came with yet another FET model. Tek themselves used many different FET models, it's not critical, so don't lose sleep over it... as long, again, as the specs the replacement you choose, meet the voltage and current rating, and take that opportunity to use a modern FET with a much lower Rdson, rather than trying to find an obsolete FET with a higher RDson... "original" is not always better ;-)
HTH,
Vincent Trouilliez
|
Re: Tek 495P A54 Memory Board Repair
Where are you located? Someone on the group may live nearby. I would offer to replace it for you, but I can't get to my workbench until I finish disposing of the things my family stored in the building.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Rick Boswell <frboswell@...>
Sent: Jul 24, 2018 1:38 PM
To: [email protected]
Subject: Re: [TekScopes] Tek 495P A54 Memory Board Repair
The switches are enabled to the data bus via tri-state buffer and an enable signal that occurs just once during boot up. Challenging to trouble shoot with my limited test equipment. And, if the buffer is bad, I lack the desoldering gear necessary to change it.
|
Re: Tek 495P A54 Memory Board Repair
The switches are enabled to the data bus via tri-state buffer and an enable signal that occurs just once during boot up. Challenging to trouble shoot with my limited test equipment. And, if the buffer is bad, I lack the desoldering gear necessary to change it.
|
Re: Tek 495P A54 Memory Board Repair
I only have the user manual for that model, but the SA has to read the switches, usually through a latch that is polled at startup. That is where I would start troubleshooting.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Rick Boswell <frboswell@...>
Sent: Jul 24, 2018 12:55 PM
To: [email protected]
Subject: [TekScopes] Tek 495P A54 Memory Board Repair
I have a nice high serial number (B030xxx) 495P spectrum analyzer with a defective A54 memory board P/N 671212401. I have determined that this board is defective by swapping in a known good board from another 495P I have replaced the usual electrolytics but no joy. With the defective board in the SA, it boots up as a 492AP instead of a 495P as shown via on-screen messages. The 49x series of SAs use a common base of microcomputer code, with the model no. (and corresponding firmware) determined via DIP switches on the A54 memory board. I have checked the DIP switches with an ohmmeter and these seem to be working as expected, and are set according to the service manual. The DIP switches are enabled to the data bus during boot up to allow these switch settings to be read. It appears that something in the circuit enabling these switches to the bus is not working correctly. I have reached the limit of my test equipment and diagnostic & repair skills and am looking for someone who might be able to diagnose and repair this A54 memory board, or, looking to purchase a known good A54 board with v9.7 firmware. Any guidance much appreciated.
Rick
K8EZB
|
Re: Bringing up a Tek 555 dual-beam scope
Hey Mike, Ive done three 500 Series restorations recently - 549, 547 and a 535,not a 555 though Tried to document as much as possible below, I don't do magic smoke method, rather take a very cautious approach - time consuming, but saves from burning part. Typically 1. I clean the unit totally, and do a complete visual inspection - for damaged/shorted components, broken wiring etc. 2. Recap - I generally replace all electrolytic and Paper. Mica ones do survive - But they all Mica/Paper the same, best way to test few with a leak tester. 3. Test tubes - Shorted tubes can be a disaster - So generally I test almost all tubes and ensure they are not shorted. 4. Naked wiring - They do have a lot of naked wires, you need to verify they are not shorted to ground/chassis. I burned many tubes while restoring 549 because of this 5. Use variac & power up, I do heater first, (remove Time delay Tube) and then if all good, then put the Delay tube back to power up all DC rails. 6. Then its troubleshooting based on what is broken - HTH Regds Rajesh On Sun, Jul 22, 2018 at 10:48 AM, Michael Leslie Squires <mikes@...> wrote: Just bought a Tek 555 scope with cart and power supply. The scope
appeared to have been used until replaced by a couple of 4xx series scopes
but just shunted aside when replaced.
I owned and used a 555 40 years ago (warmed my basement quite well) but
that scope, a Frankenstein monster created by plugging the CRT unit from
one into the power supply from another, not only worked but was calibrated
when turned on. As a result I never had to do any work on the unit.
Are there any precautions I should take when powering up this unit? One
suggestion is to plug it in and turn it on; if the magic smoke escapes find
where it came from and fix it. I do have a Variac and am thinking about
building one of the devices which has a large incandescent in series with
the DUT.
I do note that the contacts on some of the plug-ins are quite dirty, so
I'm going to give the unit a good cleaning before firing it up. I will
also check the big electrolytics for signs of failure and possibly for high
ESR.
My first scope was a Tek 511; I had to rebuild the HV supply as the diodes
had failed. I currently have a 465B which is working well.
Mike Squires
wwww.siralan.org or www.smithgreensound.com
UN*X at home since 1986
-- /Rajesh
|
Re: Homemade tunnel diodes
FWIW non-linear optical methods are used to compress the length of laser light pulses. Somewhat
analogous to NLTL's, but at 10^14Hz.
Craig
toggle quoted message
Show quoted text
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Ed Breya via Groups.Io
Sent: 24 July 2018 17:32
To: [email protected]
Subject: Re: [TekScopes] Homemade tunnel diodes
Jose said:
"Not many people seem to have made DIY NLTL, let alone samplers based on it."
This is probably true, because NLTLs usually have lots of stages, so can get very big and
complicated
before you get remarkable compression. For simplicity and effectiveness, a single SRD/varactor of
the
right type, in the right circuit environment, is pretty hard to beat for straight up impulse
generation or
frequency multiplication. A whole bunch of them cascaded can form a NLTL, with its edge-
enhancement and wide BW. As always, it depends on the particulars of the application.
I don't think there's a need for a "modern" replacement for the SRD/varactor. It's a common RF
part -
not at all an obsolete technology. Nowadays there are alternatives for many applications, by using
fast
active devices capable of appropriate edge speeds.
Ed
|
Tek 495P A54 Memory Board Repair
I have a nice high serial number (B030xxx) 495P spectrum analyzer with a defective A54 memory board P/N 671212401. I have determined that this board is defective by swapping in a known good board from another 495P I have replaced the usual electrolytics but no joy. With the defective board in the SA, it boots up as a 492AP instead of a 495P as shown via on-screen messages. The 49x series of SAs use a common base of microcomputer code, with the model no. (and corresponding firmware) determined via DIP switches on the A54 memory board. I have checked the DIP switches with an ohmmeter and these seem to be working as expected, and are set according to the service manual. The DIP switches are enabled to the data bus during boot up to allow these switch settings to be read. It appears that something in the circuit enabling these switches to the bus is not working correctly. I have reached the limit of my test equipment and diagnostic & repair skills and am looking for someone who might be able to diagnose and repair this A54 memory board, or, looking to purchase a known good A54 board with v9.7 firmware. Any guidance much appreciated.
Rick
K8EZB
|
Re: Homemade tunnel diodes
John Larkin of Highland Technology has posted a lot of comments about doing this on the news:sci.electronics.design Usenet newsgroup. His company does a lot of high speed system designs for other high tech companies. You can access that group through Google Groups to contact him, or just to read his comments.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: "Ed Breya via Groups.Io" <edbreya@...>
Sent: Jul 24, 2018 12:31 PM
To: [email protected]
Subject: Re: [TekScopes] Homemade tunnel diodes
Jose said:
"Not many people seem to have made DIY NLTL, let alone samplers based on it."
This is probably true, because NLTLs usually have lots of stages, so can get very big and complicated before you get remarkable compression. For simplicity and effectiveness, a single SRD/varactor of the right type, in the right circuit environment, is pretty hard to beat for straight up impulse generation or frequency multiplication. A whole bunch of them cascaded can form a NLTL, with its edge-enhancement and wide BW. As always, it depends on the particulars of the application.
I don't think there's a need for a "modern" replacement for the SRD/varactor. It's a common RF part - not at all an obsolete technology. Nowadays there are alternatives for many applications, by using fast active devices capable of appropriate edge speeds.
Ed
|
Re: Homemade tunnel diodes
Jose said:
"Not many people seem to have made DIY NLTL, let alone samplers based on it."
This is probably true, because NLTLs usually have lots of stages, so can get very big and complicated before you get remarkable compression. For simplicity and effectiveness, a single SRD/varactor of the right type, in the right circuit environment, is pretty hard to beat for straight up impulse generation or frequency multiplication. A whole bunch of them cascaded can form a NLTL, with its edge-enhancement and wide BW. As always, it depends on the particulars of the application.
I don't think there's a need for a "modern" replacement for the SRD/varactor. It's a common RF part - not at all an obsolete technology. Nowadays there are alternatives for many applications, by using fast active devices capable of appropriate edge speeds.
Ed
|
Hi all, a cheap and broken 2215A found it's way to my home :-), so now I am in the process of getting it repaired. I used this document from H?kan that helped me a lot: The fuse would blow instantly, so I separated the main PS from the rest by removing Q9070. An inspection by the eye learned that there were several low voltage elco's leaking, C960/C962 (+8,6V) were leaking; C961/C963 (-8,6V) were leaking; C968/C970 (+5,2V) were bad; C956 (+30V) was leaking; Fortunately no real damage to the PCB. Also the Q9070 and CR907 were toast. According to the document the FET is a IRF730 (400V 5,5A 0,75Ohm 12ns) and the diode a BYD73G (400V 1A 50ns) I replaced all the capacitors first, and now the scope has trace(s) when applying 43VDC to TP940 (Pos) and TP950 (neg). Being the optimist, I replaced Q9070 with a FQPF4N90C (900V 4A 3,5Ohm 50ns) and CR907 with a PR1507 (1000V 1,5A 300ns) I harvested from a switching PS, but.... They last about 2 seconds... Based on H?kan's document I applied a test voltage to C925, and the signals at the chip U930 check out, and since I tested all surrounding components Q908, CR908, R909, CR920, I come to think it must be the speed of the Q9070 and CR907 that causes the sudden death of these replacement components. My question is: does anybody (recently) repaired a 2215A LVPS and replaced these 2 comonents? If so, what replacements were used ? The documents suggest a MTP6N55 for the FET and a BYD73G for the diode, but I would love to hear if there are other successful candidates. Pictures can be found in this album: /g/TekScopes/album?id=64919un saludo, Leo
|
Vince
Michael A. Terrell
satbeginner
Sergey Kubushyn