Input wires go parallel with output transformer wires, and this can cause positive feedback, so oscillations.
Check if you have correct wires at X1.......(If I'm looking at your picture, I think you switched wires 15V to - 100V and viceversa between power supply and main board....).
Pull out all tubes and measure each V from power supply.
Did you use ECC 83?
Also for audio signals you need to use shielded cable.
Pull out all tubes and measure each V from power supply.
Did you use ECC 83?
Also for audio signals you need to use shielded cable.
The FFT function is useless compared to a real spectrum analyzer. I have a Keysight MSOX3104T, fully loaded with software. The FFT function is something I rarely use and I have spectrum analyzers and they do the job a lot better.
Given the cost of the less expensive DSOs, I can't imagine the FFT is that useful. I find the one in my scope frustrating to use. I guess it depends on what you are use to. I would think a budget FFT spectrum analyzer would blow what you get in a scope totally away!
Given the cost of the less expensive DSOs, I can't imagine the FFT is that useful. I find the one in my scope frustrating to use. I guess it depends on what you are use to. I would think a budget FFT spectrum analyzer would blow what you get in a scope totally away!
anatech,
1. I worked for Tektronix in the spectrum analyzer division. Manufacturing, Engineering, Marketing, and finally Technical Support.
Yeah, Tektronix: Purchased a company that made oscilloscope plug in spectrum analyzers, in New York in 1964.
The original premier analog spectrum analyzer for Audio and analysis up to 5 MHz, the 7L5.
There also was a distortion analyzer, and that eventually spinoff into Audio Precision.
The military ruggedized spectrum analyzers that even HP could not produce until years later
The MTTS show in Long Beach, California, with the 2782 spectrum analyzer (I did some of the design work). The 2782 'ran circles around' the HP flagship 8566. In 1989, I was already in marketing, and at the show.
I measured 500GHz with the 2782 or 2784, with special signal source, waveguide mixer, and waveguide adapter from small to less than mm dimensions (millimeter).
For 10 years, I also supported both the Rohde & Schwarz spectrum analyzers, and Advantest spectrum analyzers.
I also supported the Sony-Tektronix real time 3066 and 3086 spectrum analyzers.
Then I supported the RSA3000 series and RSA5000 series Tektronix real time spectrum analyzers.
Our real time spectrum analyzers had digital demodulation measurements such as EVM.
They also have very wide analysis bandwidth, to do transient capture and analysis.
And I supported other products and disciplines too.
2. If you are going to use most digital oscilloscope's FFT function, you really do have to know how to use it.
I do.
3. One of my jobs was to identify the measurement limits of any given spectrum analysis, and sometimes involved "reverse engineering" the results of a poorly adjusted spectrum analyzer, FFT, etc.
1. I worked for Tektronix in the spectrum analyzer division. Manufacturing, Engineering, Marketing, and finally Technical Support.
Yeah, Tektronix: Purchased a company that made oscilloscope plug in spectrum analyzers, in New York in 1964.
The original premier analog spectrum analyzer for Audio and analysis up to 5 MHz, the 7L5.
There also was a distortion analyzer, and that eventually spinoff into Audio Precision.
The military ruggedized spectrum analyzers that even HP could not produce until years later
The MTTS show in Long Beach, California, with the 2782 spectrum analyzer (I did some of the design work). The 2782 'ran circles around' the HP flagship 8566. In 1989, I was already in marketing, and at the show.
I measured 500GHz with the 2782 or 2784, with special signal source, waveguide mixer, and waveguide adapter from small to less than mm dimensions (millimeter).
For 10 years, I also supported both the Rohde & Schwarz spectrum analyzers, and Advantest spectrum analyzers.
I also supported the Sony-Tektronix real time 3066 and 3086 spectrum analyzers.
Then I supported the RSA3000 series and RSA5000 series Tektronix real time spectrum analyzers.
Our real time spectrum analyzers had digital demodulation measurements such as EVM.
They also have very wide analysis bandwidth, to do transient capture and analysis.
And I supported other products and disciplines too.
2. If you are going to use most digital oscilloscope's FFT function, you really do have to know how to use it.
I do.
3. One of my jobs was to identify the measurement limits of any given spectrum analysis, and sometimes involved "reverse engineering" the results of a poorly adjusted spectrum analyzer, FFT, etc.
Last edited:
Hi 6A3,
I'm not challenging your experience. However I use these products daily and have for decades. I have both Tek and HP, Agilent and Keysight oscilloscope products. Starting with the HP 3580A audio spec-an (pretty neat device!).
I am not impressed with the 2465B I have. Sorry. Terrible trace and it is slow as heck! I do have 7613 and 7633 main frames with many plug-ins, so I am familiar with those products. I will say that the myth that HP scopes don't trigger well is total bunk.
The spectrum analyzers I have do a far, far better job with spectrum analysis than the ones built into scopes. The Keysight one I use now is a $25K product and works exceptionally well. But I wouldn't ever recommend you forgo a spectrum analyzer because it does it as well. Yes, I know how to set it up.
I'm also a Journeyman calibration tech.
I'm not challenging your experience. However I use these products daily and have for decades. I have both Tek and HP, Agilent and Keysight oscilloscope products. Starting with the HP 3580A audio spec-an (pretty neat device!).
I am not impressed with the 2465B I have. Sorry. Terrible trace and it is slow as heck! I do have 7613 and 7633 main frames with many plug-ins, so I am familiar with those products. I will say that the myth that HP scopes don't trigger well is total bunk.
The spectrum analyzers I have do a far, far better job with spectrum analysis than the ones built into scopes. The Keysight one I use now is a $25K product and works exceptionally well. But I wouldn't ever recommend you forgo a spectrum analyzer because it does it as well. Yes, I know how to set it up.
I'm also a Journeyman calibration tech.
anatech,
I Never heard: "HP scopes don't trigger well".
Giving credit to ones competitors does happen.
And some companies, and some customers "bad mouth" one brand or another.
Not all products from one manufacturer are perfect.
I always enjoyed the hunt.
I used to send my customer to HP to get what he needed, if we (Tektronix) could not do what he needed to do, but HP could do it.
My customers came back to me the next time, to see what we had for his new measurement problems.
I got some of my most important electronics learning on a US Naval Destroyer in the middle of the Pacific, when we often did not have the proper parts to fix a lot of failed electronic devices.
When I think of calibration techs, I think of Luthiers.
I Never heard: "HP scopes don't trigger well".
Giving credit to ones competitors does happen.
And some companies, and some customers "bad mouth" one brand or another.
Not all products from one manufacturer are perfect.
I always enjoyed the hunt.
I used to send my customer to HP to get what he needed, if we (Tektronix) could not do what he needed to do, but HP could do it.
My customers came back to me the next time, to see what we had for his new measurement problems.
I got some of my most important electronics learning on a US Naval Destroyer in the middle of the Pacific, when we often did not have the proper parts to fix a lot of failed electronic devices.
When I think of calibration techs, I think of Luthiers.
Well, I own and use both brands. I'm not knocking one excessively.
I was very tired of hearing from many quarters that HP scopes didn't trigger well. I'm surprised you never ran into that. Completely agree no one manufacturer makes perfect products. One thing I will say is that today's instruments cost more than they should, and build quality is very low compared to what it was. Manufacturers also do not stand behind their products as they once did. Software (firmware) design also lacks, rushed to market. "We'll fix it in the field", I am so sick of that mind set.
My learning was being mentored, and work shops. I learned to do things correctly, period. At least I could wait for the correct part, but would stick in something close to see if that was the fault or not. Replacing with the correct one later. The one tech that taught me the most was Austrian, and I had to prove I was worth training.
Hmm, I had to pass Guide 25 and ISO testing. I did. I worked in a calibration lab staffed mostly by US military personnel.
I was very tired of hearing from many quarters that HP scopes didn't trigger well. I'm surprised you never ran into that. Completely agree no one manufacturer makes perfect products. One thing I will say is that today's instruments cost more than they should, and build quality is very low compared to what it was. Manufacturers also do not stand behind their products as they once did. Software (firmware) design also lacks, rushed to market. "We'll fix it in the field", I am so sick of that mind set.
My learning was being mentored, and work shops. I learned to do things correctly, period. At least I could wait for the correct part, but would stick in something close to see if that was the fault or not. Replacing with the correct one later. The one tech that taught me the most was Austrian, and I had to prove I was worth training.
Hmm, I had to pass Guide 25 and ISO testing. I did. I worked in a calibration lab staffed mostly by US military personnel.
anatech,
I went on some training trips, to teach spectrum analyzer operation, interpretation, and related subjects.
I trained US Air Force and US Air Guard, and some electronic countermeasure maintenance crews.
And I also trained 15 engineers at Tel Mex in Mexico city before it was privatized.
I went to various trade shows.
Probably my favorite was 1989 MTTS, when we introduced the 2782.
One night after the day show, HP invited us to their booth for a show and tell and questions.
Then we invited HP to our booth for a show and tell and questions. I talked to one of the major engineers of the HP8566,
and another engineer who asked if our SA 3Hz Resolution bandwidth filter was digital or analog; I said analog (he walked away scratching his head).
We were showing the 60Hz sidebands that were on a 30GHz signal source. Fun!
I have to apologize for all the readers of this thread, I went completely off the rails.
I went on some training trips, to teach spectrum analyzer operation, interpretation, and related subjects.
I trained US Air Force and US Air Guard, and some electronic countermeasure maintenance crews.
And I also trained 15 engineers at Tel Mex in Mexico city before it was privatized.
I went to various trade shows.
Probably my favorite was 1989 MTTS, when we introduced the 2782.
One night after the day show, HP invited us to their booth for a show and tell and questions.
Then we invited HP to our booth for a show and tell and questions. I talked to one of the major engineers of the HP8566,
and another engineer who asked if our SA 3Hz Resolution bandwidth filter was digital or analog; I said analog (he walked away scratching his head).
We were showing the 60Hz sidebands that were on a 30GHz signal source. Fun!
I have to apologize for all the readers of this thread, I went completely off the rails.
Hi @6A3sUMMER ,
I don't know which section of the forum is the right one, but it would be interesting to read old experiences like these for everyone won't have the possibility to learn that way.
Hi 6A3,Yes, we are off in the weeds here.
HP did design with analogue filters often, they were 1 Hz on the 3580A and 3 Hz on the 3585A. No idea why the guy walked away. Who knows? But HP went digital earlier than most and maybe he was of that new school of thinking. This is old, very reliable engineering (analogue).
A section for war stories would be interesting. Not real "war", but tales from the actual working conditions we learned under and experienced.
HP did design with analogue filters often, they were 1 Hz on the 3580A and 3 Hz on the 3585A. No idea why the guy walked away. Who knows? But HP went digital earlier than most and maybe he was of that new school of thinking. This is old, very reliable engineering (analogue).
A section for war stories would be interesting. Not real "war", but tales from the actual working conditions we learned under and experienced.
I have enjoyed operating an HP3585. Have you ever been measuring for 20 or 30 minutes, and suddenly, the instrument halts, and it calibrates those narrow resolution bandwidth filters? Then it comes back to measuring again.
Years later, HP went to using digital resolution bandwidth filters.
And so did Tektronix, Rhode & Schwarz, Advantest, etc.
Correct, this is not a "war" of who or what is best.
I had fun competing with my counterparts in other T&M companies.
I am glad for posters on this forum such as JHstewart9.
And for posters that worked for companies that were customers of HP and Tektronix, etc. I can't think of his name, but one poster worked for Motorola on RF amplifiers. He really knows how to extract the most from an output vacuum tube, and then he blows it up so you do not have to.
i am just getting too old.
Years later, HP went to using digital resolution bandwidth filters.
And so did Tektronix, Rhode & Schwarz, Advantest, etc.
Correct, this is not a "war" of who or what is best.
I had fun competing with my counterparts in other T&M companies.
I am glad for posters on this forum such as JHstewart9.
And for posters that worked for companies that were customers of HP and Tektronix, etc. I can't think of his name, but one poster worked for Motorola on RF amplifiers. He really knows how to extract the most from an output vacuum tube, and then he blows it up so you do not have to.
i am just getting too old.
Hi 6A3,
Oh gawd! We are all too old! We lived through the best times I think, the most exciting in T&M, and simply life. Seen the worst things and the best. Quite the ride in every aspect of our lives.
Yes, I still have a pair of 3585A's, one with an extremely dim tube. Yes, they stop to recalibrate every so often depending on settings. I never bothered to time it, I was busy doing work. It was very annoying! I also use a 4195A on my bench.
Yes, it was Tubelab.
Oh gawd! We are all too old! We lived through the best times I think, the most exciting in T&M, and simply life. Seen the worst things and the best. Quite the ride in every aspect of our lives.
Yes, I still have a pair of 3585A's, one with an extremely dim tube. Yes, they stop to recalibrate every so often depending on settings. I never bothered to time it, I was busy doing work. It was very annoying! I also use a 4195A on my bench.
Yes, it was Tubelab.
Yes, that would be me. I believe member rsavas also worked for Motorola in Canada, Toronto maybe.
I started at Motorola in Florida assembling HT-220 walkie talkies in 1973. I got into the cal lab in 1974 and stayed there until 1984 before going to engineering. For those 10 years I was a Mr Fixit for the production lines with regard to anything related to test and measurement. During those years virtually every piece of test equipment was HP, except for scopes, they were all TEK until we started using Hitachi scopes in the factory due to their low cost and ability to withstand the abuse handed out by factory workers.
The HP3585A and HP8903A along with a lot of their other stuff from the same time period used a proprietary HP IC process known as SOS, Silicon on Saphire for the microprocessor chip. Those chips were only made for a few years and are unobtanium today. Some have died including the one in my HP3585A which I wound up selling for parts. My HP8903A still works, probably because I have a spare CPU board. The "B" versions of these instruments have a different CPU, which is also somewhat hard to find.
There was a particular HP scope where the switch wafers were mounted on vertical PC boards with long shafts running through the boards. Whoever engineered that cluster obviously never thought about fixing it. Do switches wear out with someone turning the knobs for 3 shifts a day 24/7. That killed the HP scopes in the factory. Spending 10 years pushing a "crash cart" around a factory is a huge learning experience expecially as that factory morphed from 100% manual labor (people stuffing through hole boards and hand assembling radios) to near full automation (load up the machines, push the go button and watch populated PC boards come out the other end).
I left the cal lab in 1984 to learn about how to design radios.....radios that can be taken apart and fixed. That world changed a lot in the 30 years that I was in the middle of it. Our group morphed into an IC design center, which was above my skill set, so I retired in 2014.
Tubelab_com.
I am always glad to read your posts!
At the end of the day-shift in a secure area, the workers would twist all the knobs on the multiple spectrum analyzers, and leave.
After hours, the cleaning lady would come into the secure area to clean.
She would note all the finger marks and wear marks around the knobs of the HP 141T spectrum analyzers, to discover the usual settings.
Then, she would turn the frequency dial end to end to find the flat spot in the tuning mechanism, and write down the frequency(s).
Then she would take a picture of the CRT burn marks.
I would get calls about our Tektronix spectrum analyzers, and how to clean their memories.
Then I would tell them about the 141T's that they had.
:^)
I am always glad to read your posts!
At the end of the day-shift in a secure area, the workers would twist all the knobs on the multiple spectrum analyzers, and leave.
After hours, the cleaning lady would come into the secure area to clean.
She would note all the finger marks and wear marks around the knobs of the HP 141T spectrum analyzers, to discover the usual settings.
Then, she would turn the frequency dial end to end to find the flat spot in the tuning mechanism, and write down the frequency(s).
Then she would take a picture of the CRT burn marks.
I would get calls about our Tektronix spectrum analyzers, and how to clean their memories.
Then I would tell them about the 141T's that they had.
:^)
Yes, they also did that with access pads and a number of other things.
Now we have secure memory wipes in uP controlled instruments. Many times you pop out the HD (now SSD) and leave it there.
Now we have secure memory wipes in uP controlled instruments. Many times you pop out the HD (now SSD) and leave it there.
As you suspected, there's a problem with the grid bias:
-94V at end of disconnected lead from power supply
-68V at R28 (power tubes not installed)
-25V on V2 pin 5 (power tubes not installed)
-18V on V2 pin 5 (power tubes installed, timer relay off)
-1.3V on V2 pin 5 (power tubes installed, timer relay on)
Other odd readings without power tubes installed:
V2 pin 3 = 0V, V2 pin 4 = 403V
V3 pin 3 = 410V, V3 pin 4 = 0V
Dom
Hi Dom,
Well, there you go! Your answer is in there. Basic electronic troubleshooting ought to get you there.
Tubes normally do not draw current from their grids with negative bias wrt the cathode. You may have gassy tubes. The getter one the glass should be silver. If it is white, you have a gassy tube(s).
Well, there you go! Your answer is in there. Basic electronic troubleshooting ought to get you there.
Tubes normally do not draw current from their grids with negative bias wrt the cathode. You may have gassy tubes. The getter one the glass should be silver. If it is white, you have a gassy tube(s).
One step forward and another step back.
The tubes are good - at least they play fine in another amp. The good news is I have managed to adjust the tube bias, measuring 400mV in both EL34s for now.
I connected the amp up to a speaker, but the bad news is the music is completely drowned out by a LOUD mains hum. The hum is independent of input volume from the preamplifier and stops the second I switch off the power, with music continuing sweetly for a few seconds until the amp dies.
I have swapped out the baby Huey board, the power board and all the tubes to see if I can isolate the issue, but no joy so far. So I’m looking for any suggestions again please.
Measuring V2 and V3 pins without power tubes installed I read Pin 3 = 405V, Pin 4 = 408V, Pin 5 = -32V.
You might have noticed from the photos that I put an XLR input via a sowter 3575 1:1 input transformer. It might be a red herring, but I wonder if I’m getting some ground loops issues because I tied both the core and centre tap (pins 2 and 5 on the 3575) to chassis ground?
Any other suggestions welcome.
Dom
The tubes are good - at least they play fine in another amp. The good news is I have managed to adjust the tube bias, measuring 400mV in both EL34s for now.
I connected the amp up to a speaker, but the bad news is the music is completely drowned out by a LOUD mains hum. The hum is independent of input volume from the preamplifier and stops the second I switch off the power, with music continuing sweetly for a few seconds until the amp dies.
I have swapped out the baby Huey board, the power board and all the tubes to see if I can isolate the issue, but no joy so far. So I’m looking for any suggestions again please.
Measuring V2 and V3 pins without power tubes installed I read Pin 3 = 405V, Pin 4 = 408V, Pin 5 = -32V.
You might have noticed from the photos that I put an XLR input via a sowter 3575 1:1 input transformer. It might be a red herring, but I wonder if I’m getting some ground loops issues because I tied both the core and centre tap (pins 2 and 5 on the 3575) to chassis ground?
Any other suggestions welcome.
Dom