Hi All,

This is a follow-up post of the one called “TR7 suddenly lost power”.
To summarize, the PA was fixed, and in CW it generates 150W on 80 and 40 mts, 140W on 20, 40W on 15 and less than 10W on 10 meters, on all modes.

I checked the pin diodes in the HPF, front and rear boards, finding no open or shorted diodes.
The only diode that calls my attention is CR1501, with 600mV on the anode with the equipment on 10mts, receive condition. Based on what I have read, it should not be above 300mV, so I need to replace the diode. As I do not have UM9401s, I understand 1N4007s can be used instead. Is it so?

Something else: there is quite a difference in the SSB and CW output power on 15 and 10 mts. On 15mts, I get 90W on SSB and 40W on CW.? At the beginning, I couldn’t get more than 60W on 40m CW until I adjusted the frequency of the 5.645 KHz CW oscillator to produce a lower pitch sidetone… The maximum power happens when the pitch is close to zero Hz. I left it at 300 HZ so I can hear the sidetone. I aligned the 2nd Mixer and got a bit more power, not as much as I expected.

Any ideas?? Thanks in advance.

73 de Raoul CE3RY

300mV is a maximum tolerable voltage for an OFF condition. If you are seeing 600mV there it might be a metering issue, or CR1501 might be What is leaky. What is more important is how the diode performs in transmit. In RX there is no voltage applied to the anode so the only DC circuit path is from the junction of the PIN diodes through CR1501 and your meter to ground. If you disconnect +13V RX from the HPF module does the 600mV go away? With +13.6 RX disconnected you could short across CR1501 and see what that does for your output. You can certainly try a 1N7007, they have been used successfully in a number of RF applications.

Power out on 15 and 10 meters will always be lower by design. The gain of the transistors on the PA deck decreases as you go up in frequency. Because of this the ALC threshold for the two high band ranges is lowered by a pair of switched resistors connected to the band switch (S1401 Front on the HPF module). With a healthy PA deck and the ALC pot set for 140-150 Watts output per the service manual you should see 100 Watts, give or take a little, on 15 and 10 Meters. If you turn the rig down to 120 watts the high bands output will me proportionally less.

You may have to experiment with the tuning of T1101 to get a happy balance between CW waveform shape and power output. You might also check L1006, L1009, L1014, L1013, L1011 and L1012 on the PBT/Ref board. And of course check every point that is grounded with a screw.

73

-Jim
NU0C




On Sun, 19 Nov 2023 21:19:46 -0800
"Raoul Meunier" <rmeunier@...> wrote:

Hi All, This is a follow-up post of the one called “TR7 suddenly lost power”.
To summarize, the PA was fixed, and in CW it generates 150W on 80 and 40 mts, 140W on 20, 40W on 15 and less than 10W on 10 meters, on all modes. I checked the pin diodes in the HPF, front and rear boards, finding no open or shorted diodes.
The only diode that calls my attention is CR1501, with 600mV on the anode with the equipment on 10mts, receive condition. Based on what I have read, it should not be above 300mV, so I need to replace the diode. As I do not have UM9401s, I understand 1N4007s can be used instead. Is it so? Something else: there is quite a difference in the SSB and CW output power on 15 and 10 mts. On 15mts, I get 90W on SSB and 40W on CW. At the beginning, I couldn’t get more than 60W on 40m CW until I adjusted the frequency of the 5.645 KHz CW oscillator to produce a lower pitch sidetone… The maximum power happens when the pitch is close to zero Hz. I left it at 300 HZ so I can hear the sidetone. I aligned the 2nd Mixer and got a bit more power, not as much as I expected.

Any ideas?? Thanks in advance.

73 de Raoul CE3RY

Thank you very much, Jim. I will try your suggestions and I'll comeback.
Just for trial purposes, I swapped the 2nd IF and the 2nd Mixer boards from my "working TR7"... The output increased to almost 100W on 15m but remained at 10W on 10m. I did not touch anything in the swapped boards, though.
There are some component differences between the two 2nd Mixer boards. The "best" one has everything by the book. The other one seems to be an earlier component version of which I have no diagram for it. I believe I can modify it to become the newer version.

Thanks again and 73,
Raoul CE3RY

Jim,?

Some comments:

Q - "If you disconnect +13V RX from the HPF module does the 600mV go away?"
A - Yes, I read zero volts in that condition.

Q - With +13.6 RX disconnected you could short across CR1501 and see what that does for your output.
A - Nothing different happens, same result as before.

Q - You can certainly try a 1N4007...
A - With the 1N4007 I read almost zero volts in Rx condition. I can see C1501 discharging. BTW, this capacitor was changed for a .47uF. I'll replace it with a .22 uF per the diagram.

Q - You may have to experiment with the tuning of T1101... etc... You might also check L1006, L1009... etc
A - I did all of that and only got a few watts more on 15 and 10 meters.

I'll keep on searching!
Thanks and 73,

Raoul CE3RY

There were quite a few changes to the 2nd Mixer. I identified many of them by compairng boards and pictures harvested from the internet, and Peter (VE7PS) sent what should be a complete list.

---
From: "VE3AX/7" <VE3AX@...>
Subject: Re: TR-7 Version 2 Mixer Board
Date: Thu, 29 May 2014 16:37:40 -0700

Jim (and Garey)

While I have yet to pull the TR-4310 from the rack desk, I did get a chance
to spend a bit of time today with the TR-7 and TR-4310 Service Manuals to
compare the 2nd Mixer boards. Here are my findings;

Both the TR-4310 and TR-7 pictures are of a board with the nomenclature of
C3658W, but they are clearly different.

For the TR-4310......

At Q701 - R703 is changed to a 1k (2.2k in the TR-7)
R704 is changed to a 22 ohm (33 ohms in the TR-7)
R702 (2.2k) and C 703 (.01) are removed.
T701 is a tapped coil rather than a transformer.
C737, a 10uf tantalum or electrolytic is added in
parallel with C706 to ground. On the annotated TR-7 schematic that Garey
has, this cap is written in, but is in parallel with
C702, not C706.
C704 (390) and R706 (5.1) are removed.


At T702 - R 731 (8.2k) is removed. Surprisingly, the annotated TR-7
manual has a note stating "missing? self osc" which implies to me that
one would want this resistor left in there. Hmmmm!

At Q702- C717 is changed to a .22 uf tantalum/electrolytic (.001 in the
TR-7)

On the balanced mixer diode strings, the two diodes in each string are
replaced with one diode and a 15 ohm resistor. Diodes are still 1N4148's.

A 4A6 ferrite bead is added on the line from the center tap of the T704
primary to C725.

C736, a 51 pf cap, is added to ground at the juncture of CR705 and CR706.

A U-shaped ground tab has been added to the lower edge of the board (left
side in the pic), immediately below the 7 pin edge connector. This would
rest against the chassis when the board is installed.


I think that's pretty well all, though I may have missed something. I will
scan these next Monday when I am back at work (just working 3 days/week and
retiring June 30th, supposedly) and send these off to you both. In the
m,eantime \i will try to get the 4310 out of the rack this weekend and get
some pics as well.

73 Peter
VE3AX/7
---

I have implemented most of these changes successfully. I don't recall what the state of T701 was.

Have you taken a look at what the input levels to the predriver board are compared to the measured values in K8AC's article on rebuilding the PA deck?

73




On Mon, 20 Nov 2023 11:53:51 -0800
"Raoul Meunier" <rmeunier@...> wrote:

Thank you very much, Jim. I will try your suggestions and I'll comeback.
Just for trial purposes, I swapped the 2nd IF and the 2nd Mixer boards from my "working TR7"... The output increased to almost 100W on 15m but remained at 10W on 10m. I did not touch anything in the swapped boards, though.
There are some component differences between the two 2nd Mixer boards. The "best" one has everything by the book. The other one seems to be an earlier component version of which I have no diagram for it. I believe I can modify it to become the newer version.

Thanks again and 73,
Raoul CE3RY

Comments inline below.

On Mon, 20 Nov 2023 15:01:15 -0800
"Raoul Meunier" <rmeunier@...> wrote:

Q - " If you disconnect +13V RX from the HPF module does the 600mV go away?"
A - Yes, I read zero volts in that condition.

That seems so suggest that the 600 mV reading is caused by leakage through CR1501 and your meter.

Q - With +13.6 RX disconnected you could short across CR1501 and see what that does for your output.
A - Nothing different happens, same result as before.

This should eliminate the possibility of any forward biased fault in CR1501. CR1503 and/or 1502 could be leaking or the problem could be elsewhere. Removing C1503 from the circuit and checking TX would tell us something.

Q - You can certainly try a 1N4007...
A - With the 1N4007 I read almost zero volts in Rx condition. I can see C1501 discharging. BTW, this capacitor was changed for a .47uF. I'll replace it with a .22 uF per the diagram.

Probably a good idea. The larger cap would hold the TX path open for a tiny bit lonter.

Q - You may have to experiment with the tuning of T1101... etc... You might also check L1006, L1009... etc
A - I did all of that and only got a few watts more on 15 and 10 meters.

I'll keep on searching!
Thanks and 73,

This is turning into quite a chase.

73

-Jim
NU0C

Jim,

I received? your message while I was writing this one, I'll read? yours after sending mine. Sorry...

- - - - - - - - - - - - -?

If I remove C1503 I’ll get no RF out to the PA… (?)? ?I thought of removing C1507 but haven’t done it yet.

Last night I thought “What if we don’t have enough drive to feed the PA…” This morning I used a 100MHz scope to measure the RF Voltage at the end of the cable coming from the Up Converter with and without a 50-Ohm load.? The instrument I used is a 100-MHz scope with a regular scope probe. ?I tested both TR7s I have, a “good” one, N° 4089 and the “problem” one, N° 2904. Please see the attached chart.

As I had to pull the cable from the HPF to measure the RF coming from the Up Converter, the table shows the maximum RF being delivered to the HPF and not necessarily correspond to the amount needed to produce the power output shown in the chart. ALC plays a role in these numbers.

As you will see, there are differences between the two rigs regarding the capacity to produce RF to feed the HPF.? At first sight, I would say #2904 is "short of mV" to produce the expected power, so I need to find the origin of this “problem”.

Your thoughts will be very welcome.

Thanks, 73 de Raoul CE3RY

Jim,

I had not read your message on the Second Mixer C3658W changes...
Wow! What a time saver it is!? It's certainly a keeper.
I believe I'll think twice before proceeding with these changes, at least for the moment.?
Thanks very, very much for the info.

73 de Raoul CE3RY

I meant remove CR1503. This would remove the RX antenna path from the equation.

Can you compare the levels of the LO and 2nd Mixer into the Upconverter on both rigs? That might be helpful. There were a few changes to the Upconverter as well. I'll see if I can find my notes. I've been trying to document as much of these things as I can find.

...found them.

Upconverter Board

R22 changed from 47 Ohm to 22 Ohm.
C428 10 uF Tantalum added.
L402 changed from 7 turns to 8 turns, parallel 82 pF capacitor removed.
C409 and C429 replaced with single 160 pF silver mica at C409.
22 Ohm from pins 3 & 4 of U401 changed to 4A6 ferrite bead (L407).
RFC401, RFC403, RFC407, RFC408 changed to VK-200-2 ferrite EMI chokes.


Also, the 48.05 MHz roofing filter circuit needs to be sweep aligned. If someone messed with the tuning that could cause problems. Probably not likely but worth keeping in mind. Some of the 2-pole crystal filters seem to failing or going out of tolerance as they age. I have seen a couple myself that are asymmetrical and it can't be completely tuned out. No source of replacements has been found as of yet, and some of us have looked. If I knew anything about crystal filters I would attempt to design a modern replacement.


73

-Jim
NU0C



On Tue, 21 Nov 2023 11:45:45 -0800
"Raoul Meunier" <rmeunier@...> wrote:

Jim,

I received? your message while I was writing this one, I'll read? yours after sending mine. Sorry...

- - - - - - - - - - - - - If I remove C1503 I’ll get no RF out to the PA… (?) I thought of removing C1507 but haven’t done it yet. Last night I thought “What if we don’t have enough drive to feed the PA…” This morning I used a 100MHz scope to measure the RF Voltage at the end of the cable coming from the Up Converter with and without a 50-Ohm load. The instrument I used is a 100-MHz scope with a regular scope probe. I tested both TR7s I have, a “good” one, N° 4089 and the “problem” one, N° 2904. Please see the attached chart. As I had to pull the cable from the HPF to measure the RF coming from the Up Converter, the table shows the maximum RF being delivered to the HPF and not necessarily correspond to the amount needed to produce the power output shown in the chart. ALC plays a role in these numbers. As you will see, there are differences between the two rigs regarding the capacity to produce RF to feed the HPF.? At first sight, I would say #2904 is "short of mV" to produce the expected power, so I need to find the origin of this “problem”.

Your thoughts will be very welcome.

Thanks, 73 de Raoul CE3RY

Jim,

I did remove CR1503, same behavior.

"Can you compare the levels of the LO and 2nd Mixer into the Upconverter on both rigs?" - I will try that tomorrow.

Thanks for the Up Converter changes...
One thing I noticed in this TR7 is that somebody replaced most Tantalum capacitors with electrolytic caps. Unfortunately not necessarily of the same capacity...

Thanks and 73 de Raoul CE3RY

Yeah, that could be a problem. There is absolutely no reason to replace perfectly good tantalums. Properly used they are more reliable than electrolytics. Yes, they do fail occasionally, and occasionally fail spectacularly. But they don't leak, change value much, or dry out and increase in ESR. Changing a value in controlled circuits can affect change over timing or carefully crafted R/C time constants. Not good.

73

-Jim
NU9C

On Tue, 21 Nov 2023 19:58:52 -0800
"Raoul Meunier" <rmeunier@...> wrote:

One thing I noticed in this TR7 is that somebody replaced most Tantalum capacitors with electrolytic caps. Unfortunately not necessarily of the same capacity..

Jim,

Thks for your comment on Tantalums...?

I have not yet compared the levels of the LO and 2^nd mixer into the Up Converter…

Instead, I swapped all boards -but the power supply one- from TR7s #4089 and #2904.? I know the ones that were installed in the newer TR7 work well, so I was curious to find out what would come from this test.

I found that the old boards produce enough RF to excite the PA and get 100W on 10m, 110W on 15m and 140W on 20, 40 and 80 meters. I only had to increase the pot in the Predriver to attain the 100W?on 10m.

With the newer boards in the old rig, I get 25W on 10m, 100W on 15m and 140 on 20, 40 and 80 meters. Knowing there’s enough RF coming from the Up Converter, I checked the RF at the output of the HPF, the line feeding the PA… about 15% of the input. I’ll get in the HPF later today or tomorrow…
I know, I should have done this long time ago…

73 de Raoul CE3RY

Hi All,

It's been a while since my last post on this subject. This, due to traveling...

I found the HPF? is in good conditions. Still, I can only get 50W on 15 and 10W on 10 meters.

There are a few things I haven't done yet, like replacing the electrolytics? with tantalums, as described in the manual, so I haven't finished my homework.

I found a couple electrolytics on the bottom of the parent board, between the +10V line and ground and the +5V line and ground. These are 22uF and 11uF respectively. My other TR7 has tantalums in those places. I have not found these in the manual so I wonder about their value... (please help...).

Thanks and 73,

Raoul CE3RY

Raul,

with you tests and measures I think you have nailed down the issue to the PA brick and LPF board. As such you repaired the PA before tou need to fully test it, it is the first suspect. Please read the link below, you really need to measure voltage input and out of the PA, as done in the reference, to assess gain at various bands and particularly its ability of producing power at 10m.

Second candidate for measures is LPF and relative band switching, do not trust at? measures at low power.

73

Giuliano, IV3DLW?

Ps: I understand that in your table, power output at PA means power output measured at the PL258 connector of RTX, AFTER the LPF.
Giuliano, IV3DLW?

Hi Giuliano,

Thanks for your messages. You are right; in the table I sent, power output means the power measured after the SO239 connector on the back of the TRx and, naturally, after the LPF.
As you point it out, the LPF was a suspect, so I disconnected the coax going to the LPF and used a short piece of coax to a Dummy Load. On 15 meters, the reading increased five watts. On 10 mts I noticed only a couple watts more. All of these measurements in CW mode, key down.

So I tried SSB... On 15 mts I got peaks of 90 Watts measured with an Autek Research WM-1 Computing Meter. On 10 mts, the power output was doubled, up to 18 Watts.
So, I think there is something else that I have not found yet for I shouldn't be getting that huge difference in output power when in CW and SSB modes.
Any suggestion is welcome!

Thanks and 73 de Raoul CE3RY

?

Keep in mind that the CW carrier comes from a different source than the SSB modulation. AM carrier re-insertion is different again and may not reach the same level.

73

-Jim
NU0C

On Fri, 22 Dec 2023 11:58:06 -0800
"Raoul Meunier" <rmeunier@...> wrote:

Hi Giuliano,

Thanks for your messages. You are right; in the table I sent, power output means the power measured after the SO239 connector on the back of the TRx and, naturally, after the LPF.
As you point it out, the LPF was a suspect, so I disconnected the coax going to the LPF and used a short piece of coax to a Dummy Load. On 15 meters, the reading increased five watts. On 10 mts I noticed only a couple watts more. All of these measurements in CW mode, key down.

So I tried SSB... On 15 mts I got peaks of 90 Watts measured with an Autek Research WM-1 Computing Meter. On 10 mts, the power output was doubled, up to 18 Watts.
So, I think there is something else that I have not found yet for I shouldn't be getting that huge difference in output power when in CW and SSB modes.
Any suggestion is welcome!

Thanks and 73 de Raoul CE3RY

--

73

-Jim
NU0C

Thanks Jim, you are right.? But, why is it that it's only different on 15 and 10 mts only? On the other bands, CW and peaks of SSB reach the same level... AM is a different story.

Thanks and 73,?
Raoul CE3RY

Raoul,

Only meaures will tell where the issue really is.

The link I provided points out the fact that a partial failure of RF transistors (especially the drivers) could result in severe power shortcut at higher frequencies. We need an explanation of this fact, because, at some extent, it is counter-intuitive.
We are usually thinking that a solid state device is good or gone, and this is normally true, but this is not the case.
RF power? transistors (finals and drivers) are in fact special devices, constituted internally by a certain number of transistors in parallel. How many it depends from the specific design.?
We must also consider that transistor's gain get down with frequency increase.?
So, in a wideband 100W RF system with standard MRF power transistor you need only a fraction of watt input at 3Mhz to fully drive the finals but you must provide about 5 W at 30 Mhz, for full power output.
If the majority of the small transistors inside the driver(s) are gone open (the usual case),? tests tell the device is good but in reality the driver is capable to deliver only a reduced power to the finals, proportional to the number of transistors still alive, inside it. That power should suffice for finals? to still develop full power in low bands but cannot let the finals to produce full power out, at higher frequencies, because of lack of input power.?

That said, it does not means you have the drivers toasted. Can be also other things. LPF, improper ALC, protections? (associated? power out and swr sensors), switches etc...

Vr

Giuliano, IV3DLW?

开云体育