开云体育

Jim,

Thanks for your continuing advice and help.

I have had boards out and given each contact in the female Molex connector shells of the translator and VCO boards a spot of DeoxIT D100L and scrubbed them through with a pipe cleaner.

I have cleaned the mating pins of the parent and digital display boards with WD-40 Specialist Contact Cleaner, which I have been looking to try for a while as an adjunct / alternative to DeoxIT, which is expensive / hard to source on this side of the pond. I've also re-flowed the solder on the pins of the large Molex on the display board.

No sign of any loop unlocking now. If it stays that way, I will move on to look at the transmitter low power output on 10 metre band, which is acute.

Regards,

Mark, G4FPH.

Synth loop still locked and solid:-) Started to investigate the transmitter low power issue.

Scoping the link between the pre-driver and driver and PA boards, I'm seeing the same situation as Stefan, DL7MAJ, documented in his excellent YouTube video (3/5). With the link broken, there is plenty of clean output signal from the pre-driver on all bands (so replacement of UM9401 PIN diodes with 1N4007 looks good). With the link connected, the level is lower (as expected) and the waveform is heavily distorted on many bands.

My guess is that one of the driver transistors is not pulling its weight / is broken, so I start a search for suitable replacements. The manual shows 'SRF2388', which I can't find any data on right now. The one I can see in my brick is Motorola and look original - marked 'A4018' 'M005'.

I see some have used 2SC1969 (assume these need an insulating washer as their tab is collector, same as originals?). Sadly, I have only one of those in my junk box and assume all others available to buy will be Chinese fakes. An ambitious replacement would be the RD16HHF1 RF Power MOSFET. Easier to find genuine ones of these, but not sure I have the hours to put in to the R&D.

Regards,

Mark, G4FPH.

The 2SC1969 is an acceptable sub if you can find them. RFParts used to have them in matched pairs but apparently not anymore. They can still be found but you know about the Chinese counterfeits. Any of the common final transistors used in (legal) SSB CB radios should work. You can do a quick check of the drivers by *carefully* measuring the DC base voltage. It should be very nears zero in RX and arounf 0.6V in TX with no drive. It won't guarantee both are good but it will show up leakage or a B-E short.

See also

73

-Jim
NU0C

On Mon, 19 Feb 2024 02:44:41 -0800
"atlasstuff" <g4fph@...> wrote:

Synth loop still locked and solid:-) Started to investigate the transmitter low power issue.

Scoping the link between the pre-driver and driver and PA boards, I'm seeing the same situation as Stefan, DL7MAJ, documented in his excellent YouTube video (3/5). With the link broken, there is plenty of clean output signal from the pre-driver on all bands (so replacement of UM9401 PIN diodes with 1N4007 looks good). With the link connected, the level is lower (as expected) and the waveform is heavily distorted on many bands.

My guess is that one of the driver transistors is not pulling its weight / is broken, so I start a search for suitable replacements. The manual shows 'SRF2388', which I can't find any data on right now. The one I can see in my brick is Motorola and look original - marked 'A4018' 'M005'.

I see some have used 2SC1969 (assume these need an insulating washer as their tab is collector, same as originals?). Sadly, I have only one of those in my junk box and assume all others available to buy will be Chinese fakes. An ambitious replacement would be the RD16HHF1 RF Power MOSFET. Easier to find genuine ones of these, but not sure I have the hours to put in to the R&D.

Regards,

Mark, G4FPH.

Jim,

Further positive progress here. The two driver transistors that were installed had 160 mV DC on their bases when on receive, rather than zero, so suspect leaky. I measured their standing current at the 'test jumper'. On transmit with no drive, they were pulling around 900 mA, which seemed excessive, assuming they are operated in class AB.

I took both out and tested them with my multi-meter and component tester. Transistor #1 was leaky B-C (1.7 kOhm) and showed an hFE of only 4 at a few mA. Transistor #2 B-E and B-C junctions looked to be intact, but showed 300 kOhm between C and E. The transistor tester awarded it an hFE of 40 though. So both clearly faulty, but for different reasons.

Scratching my head regarding close-at-hand replacements, I suddenly remembered I had an old Cubic Astro 102 BXA in storage that I bought as a challenge. I did get it working, with the exception of the PA brick. The four flange mount Motorola devices in that had been removed by a previous owner and, in place of the final transistors, a pair of (yes!) 2SC1969 had been hacked in. With nothing to lose, I removed the 2SC1969, cleaned up their legs and tested them. Bingo! Two good, original manufacturer, devices and very closely matched, with hFE of 45-ish.

I fitted them to the TR-7 and checked the standing current. Down to 600 mA, which still seems like a hammering to me given their position. I would not expect the final devices to be biased that hard. What are others running a pair of these at?

RF waveform on the pre-driver link is much improved, although at frequencies less than 7 MHz, there's still quite a bit of distortion. Running CW, if I turn the carrier control way down, the distortion looks more like an HF oscillation superimposed on the sin wave, which I shall look into.

Power output-wise, with max. set to 120 W on 14 MHz, I can now get up to 80 W on 28 MHz, depending on the setting of the gain adjust pot on the pre-driver board. If it proves stable at that setting, I'll be happy.

I hope this saga has been interesting to group members'. I am expecting it to end soon and that I shall be able to put the covers on and maybe even have a contact, or two. Standing back and considering the stock faults the TR-7 suffers from, I don't think any new ground has been broken.

Regards,

Mark, G4FPH.

Great! It seems strange that the drivers can fail in ways that are not a complete failure. Been there, done that. I agree that 600 mA is too much. These transistors are typically biased around 70 or 80 mA for a single transistor in CB final service. Garey K4OAH (SK) origninally brought this up and I always add a second 47 Ohm resistor in series with R2303 per the article by Floyd K8AC. That should bring it down to 200-ish mA.

For stability, ground the upper right rear corner of the High Pass Filter module as shown in the image attached here:

/g/DRAKE-RADIO/message/58910

Your power output sounds about typical. I usually go for 140-ish watts on 20m and see 100 to 110 out on 17m and above.

73

-Jim
NU0C


On Mon, 19 Feb 2024 15:14:44 -0800
"atlasstuff" <g4fph@...> wrote:

Jim,

Further positive progress here. The two driver transistors that were installed had 160 mV DC on their bases when on receive, rather than zero, so suspect leaky. I measured their standing current at the 'test jumper'. On transmit with no drive, they were pulling around 900 mA, which seemed excessive, assuming they are operated in class AB.

I took both out and tested them with my multi-meter and component tester. Transistor #1 was leaky B-C (1.7 kOhm) and showed an hFE of only 4 at a few mA. Transistor #2 B-E and B-C junctions looked to be intact, but showed 300 kOhm between C and E. The transistor tester awarded it an hFE of 40 though. So both clearly faulty, but for different reasons.

Scratching my head regarding close-at-hand replacements, I suddenly remembered I had an old Cubic Astro 102 BXA in storage that I bought as a challenge. I did get it working, with the exception of the PA brick. The four flange mount Motorola devices in that had been removed by a previous owner and, in place of the final transistors, a pair of (yes!) 2SC1969 had been hacked in. With nothing to lose, I removed the 2SC1969, cleaned up their legs and tested them. Bingo! Two good, original manufacturer, devices and very closely matched, with hFE of 45-ish.

I fitted them to the TR-7 and checked the standing current. Down to 600 mA, which still seems like a hammering to me given their position. I would not expect the final devices to be biased that hard. What are others running a pair of these at?

RF waveform on the pre-driver link is much improved, although at frequencies less than 7 MHz, there's still quite a bit of distortion. Running CW, if I turn the carrier control way down, the distortion looks more like an HF oscillation superimposed on the sin wave, which I shall look into.

Power output-wise, with max. set to 120 W on 14 MHz, I can now get up to 80 W on 28 MHz, depending on the setting of the gain adjust pot on the pre-driver board. If it proves stable at that setting, I'll be happy.

I hope this saga has been interesting to group members'. I am expecting it to end soon and that I shall be able to put the covers on and maybe even have a contact, or two. Standing back and considering the stock faults the TR-7 suffers from, I don't think any new ground has been broken.

Regards,

Mark, G4FPH.

--

73

-Jim
NU0C

Jim,

I added a 68 R in series with the existing R2303 (which was 47 R). That brought the driver standing current down to about 220 mA, which seems reasonable to me and is more in line with your note. At max RF out, the driver DC current peaks to around 700 mA, so nine and a bit Watt going in.

As with the rest of this TR-7, I am not the first to crack open the PA brick. The final devices are MRF421, with a 1990 date code, which is great. The driver transistors had been changed at least once prior to my putting Japanese ones in.

I did the ALC mod. and added the HPF ground strap, followed by an alignment and screwed the screening plates back in place, but not yet put the covers back on;-)

Imagine my joy to find most of HF to be covered by some new, pulsing, buzzing interference:-( Out with the DF loop tomorrow, I think. Be nice to get some fresh air after all the solder smoke - HI!

Regards,

Mark, G4FPH.

Awseome! I'm glad you got it up and running. Another prodigal child returns to the flock. I feel your pain, my local noise floor has increased dramatically in recent years. On 80m particularly, S-9+ although Clipperton was heard through it on CW. Tell us about your DF loop. :)

73

-Jim
NU0C

On Tue, 20 Feb 2024 09:20:19 -0800
"atlasstuff" <g4fph@...> wrote:

Jim,

I added a 68 R in series with the existing R2303 (which was 47 R). That brought the driver standing current down to about 220 mA, which seems reasonable to me and is more in line with your note. At max RF out, the driver DC current peaks to around 700 mA, so nine and a bit Watt going in.

As with the rest of this TR-7, I am not the first to crack open the PA brick. The final devices are MRF421, with a 1990 date code, which is great. The driver transistors had been changed at least once prior to my putting Japanese ones in.

I did the ALC mod. and added the HPF ground strap, followed by an alignment and screwed the screening plates back in place, but not yet put the covers back on;-)

Imagine my joy to find most of HF to be covered by some new, pulsing, buzzing interference:-( Out with the DF loop tomorrow, I think. Be nice to get some fresh air after all the solder smoke - HI!

Regards,

Mark, G4FPH.

--

73

-Jim
NU0C

On Thu, Feb 8, 2024 at 07:08 PM, atlasstuff wrote:

The UM9401 PIN diodes had been replaced with an assortment of silicon rectifier diodes. I have some sympathy here. Access to the rear of the board in-situ to make a repair job is not easy. I removed all three diodes, cleaned the rear of the board with solder wick and re-built the three diodes in a Christmas tree on the front of the board. I did not have any of the original PIN diodes. Looking at this article I substituted 1N4007. They work well on RX, but remains to be seen whether they will be good enough on TX. Any group experiences here?

According to the 1N4007 datasheet, the diode has 30pF at 0V. That is approximately 177 ohm at 30MHz.

UM9401 has approximately 1,1pF

-

docs:


backup:
https://web.archive.org/web/20240224081724/https://www.microsemi.com/document-portal/doc_view/11327-um9400-series-datasheet


backup:
https://web.archive.org/web/20100502002341/https://www.vishay.com/docs/88503/1n4001.pdf



--

--
best regards,

Glenn, OZ1HFT

Glenn,

Yes, 1N4007 performance as PIN at 30 MHz in that regard is not great. The measurements documented on the page below suggest around 6.5 pF at 10 MHz with some modest level of reverse bias, which seems in line with the datasheet. I wonder whether this issue - even with the original PIN diodes - was the reason that Drake included two in series in the receive path?



As my junk box has no PIN diodes in it, I shall check to see what can be purchased.

Regards,

Mark, G4FPH.

Jim,

Thank goodness the interference was gone the next day and has not returned. I hope it never does. Sounded like one of these horrible Chinese, switch mode / pulsed chargers for automotive batteries. So, my loop stayed under the EMC emergency / break glass cover in the shack - HI!

Please find attached a picture of the loop. It's very simple but very effective, with deep nulls. Built on a plastic, electric fence post, to which I also attach my Kenwood TH-F6 HT as detector / receiver, and designed to be held in one hand while walking about. With no amplification, the combination is easily sensitive enough to hear signals and noises. One enhancement I have thought about is something to resolve the ambiguity of having two nulls per 360-degree of revolution. In practice, however, I have not found this essential as, by making an educated guess / walking about awhile, I have been able to work out which of the two possible directions noises come from.

I keep the loop tuned for the 80-metre band. It comprises three turns of single-strand, insulated wire at around 270 mm diameter. Tuned with the parallel combination of around 210 pF, plus the air variable, which is probably set to 100 pF for mid-band. The pick up loop is one turn of the same stiff wire at around 100 mm diameter, direct connected to a short length of coax terminated with an SMA to connect to the radio.

I commend having one, or similar, to all that try to make HF radio work in non-ideal environments. Like we used to keep spare fanbelts in the boots / trunks of our cars as a guarantee that we would never need them!

Regards,

Mark, G4FPH.

Thanks for the pic and info Mark. That looks like something I would build! I've been meaning to work up something to carry my LAN-IQ portable SDR receiver around. Yes, bearings from two or more locations, a.k.a. triangulation, is a common method of resolving the location of an RF source. I once tracked down the location of a fox in a rare 80m fox hunt here by using the null off the end of the whip antenna of a portable SW receiver while the other hunters were chasing ghosts at the other end of the county. They were all VHF hunters...

73

-Jim
NU0C

On Sat, 24 Feb 2024 02:16:17 -0800
"atlasstuff" <g4fph@...> wrote:

Jim,

Thank goodness the interference was gone the next day and has not returned. I hope it never does. Sounded like one of these horrible Chinese, switch mode / pulsed chargers for automotive batteries. So, my loop stayed under the EMC emergency / break glass cover in the shack - HI!

Please find attached a picture of the loop. It's very simple but very effective, with deep nulls. Built on a plastic, electric fence post, to which I also attach my Kenwood TH-F6 HT as detector / receiver, and designed to be held in one hand while walking about. With no amplification, the combination is easily sensitive enough to hear signals and noises. One enhancement I have thought about is something to resolve the ambiguity of having two nulls per 360-degree of revolution. In practice, however, I have not found this essential as, by making an educated guess / walking about awhile, I have been able to work out which of the two possible directions noises come from.

I keep the loop tuned for the 80-metre band. It comprises three turns of single-strand, insulated wire at around 270 mm diameter. Tuned with the parallel combination of around 210 pF, plus the air variable, which is probably set to 100 pF for mid-band. The pick up loop is one turn of the same stiff wire at around 100 mm diameter, direct connected to a short length of coax terminated with an SMA to connect to the radio.

I commend having one, or similar, to all that try to make HF radio work in non-ideal environments. Like we used to keep spare fanbelts in the boots / trunks of our cars as a guarantee that we would never need them!

Regards,

Mark, G4FPH.

--

73

-Jim
NU0C

Nice Loop. I'm planning a somewhat larger one. So far, I only have
the capacitor. 15-250pF @ 45kV.

- Jerry, KF6VB

Hi Mark

PIN diode inspiration:





Found via:


backup:


--

best regards,

Glenn, OZ1HFT

Glenn,

Thanks for the data. I will be looking out for PIN suitable for the TR-7 for my spares draw.

The three 1N4007 I substituted for the RX input / TX drive steering PIN on the HPF board are working OK, but I have made no measurements of the impact of their higher capacitance. A small relay might also make a good substitute for very expensive / unobtanium PIN. They are cheap enough, but parameters of particular interest to RF engineers are not usually specified, so it would be a case of suck it and see!

Regards,

Mark, G4FPH.

Both can still be found.







73

-Jim
NU0C

On Sat, 30 Mar 2024 01:46:08 -0700
"Glenn, OZ1HFT" <glenn.mh.dk@...> wrote:

Hi Mark

PIN diode inspiration:





Found via:


backup:


--

best regards,

Glenn, OZ1HFT

--

73

-Jim
NU0C

On 2024-03-30 01:46, Glenn, OZ1HFT wrote:

Hi Mark
PIN diode inspiration:

*** Also check the QRZ page of W6JL. The master of using 1N4007s etc as PIN diodes.

- Jerry, KF6VB