A4 board update.
Undid all the mods except the LT1468 on it.
I removed the added balance pot with the
other resistors clipped off the jumpered resistors.
Did have a question as I also changed A4U1 to the LT1468 also.
Concerned that the feedback resistor is 220K ohms which seems
really high. That also fed a 200 ohm resistor to ground which
I changed back to 1k.
Just a quick spot check, it is about 2dB worse than the
substitute board without any adjustments to the HP339a.
I'll have time to tackle the other stuff soon.
Have a good weekend y'all.
David, yep I've been thinking about building shed like a little mini
barn about 3m x 5m, 4m tall with a 2nd level to store stuff.
Insulated, heated, A/C, the like. Peace can be a good thing.
Demian, will be working on that, getting the analyzer worked out.
Undid all the mods except the LT1468 on it.
I removed the added balance pot with the
other resistors clipped off the jumpered resistors.
Did have a question as I also changed A4U1 to the LT1468 also.
Concerned that the feedback resistor is 220K ohms which seems
really high. That also fed a 200 ohm resistor to ground which
I changed back to 1k.
Just a quick spot check, it is about 2dB worse than the
substitute board without any adjustments to the HP339a.
I'll have time to tackle the other stuff soon.
Have a good weekend y'all.
David, yep I've been thinking about building shed like a little mini
barn about 3m x 5m, 4m tall with a 2nd level to store stuff.
Insulated, heated, A/C, the like. Peace can be a good thing.
Demian, will be working on that, getting the analyzer worked out.
Using an LT1468 for U1 could cause stability problems. U1 has 46dB AC gain (200k/1k) and is intended to be a clipper amp that fully drives the following diode clipper strings feeding the balanced modulators to full on, giving a clean fast rise pulse waveform to do the phase comparisons. Look at the output of U1 (with a 1kHz input signal to the analyzer) with a wideband scope to be sure that the output is not full of noise. I really don't know what theoptimum gain of U1 should be, but I think maybe the HP guys did.
RichEEM,
That is A4U1 on the removable board right?
Just checking don't take offense.
If the LT1468 is too much of a replacement
I have two other choices:
OPA134
LME49710
I can measure then check and see which works best
as I do have it socketed.
I do know this current board has low frequency problems
on it. This board has about 8 - 10 dB greater distortion
across the BW from 10Hz to 10kHz than the substitute board.
The substitute board is just about spot on from 10Hz to 1kHz and
just about 1/2 to 1 dB at 10kHz.
I want to get it working again before proceeding.
I don't want to make any changes to the substitute board
because it is a known working board. At least now I have
two boards I can reference.
I did find that one possible reason after doing the gain mod didn't work
was that I had desoldered two legs off one of the quad OpAmps.
Turning that thing over, side to side I started getting lost...
until undoing the mods did it become clear to me what had
happened.
That is another challenge of doing this stuff, working alone
and not in a lab or with colleagues present, we can create
problems that we don't see. Any one just looking
at it from the outside might find it instantly or with
little problem, but we just can't see it.
Like programming, writing, math, or other tasks etc.,
the challenge is looking at things from a fresh perspective.
We just have to leave it for a time, then come back and
investigate with a fresh perspective, keeping in mind it has
to be a fresh look, which is why sites like DIY and others
can be very helpful.
I'll pop the two in and see.
That is A4U1 on the removable board right?
Just checking don't take offense.
If the LT1468 is too much of a replacement
I have two other choices:
OPA134
LME49710
I can measure then check and see which works best
as I do have it socketed.
I do know this current board has low frequency problems
on it. This board has about 8 - 10 dB greater distortion
across the BW from 10Hz to 10kHz than the substitute board.
The substitute board is just about spot on from 10Hz to 1kHz and
just about 1/2 to 1 dB at 10kHz.
I want to get it working again before proceeding.
I don't want to make any changes to the substitute board
because it is a known working board. At least now I have
two boards I can reference.
I did find that one possible reason after doing the gain mod didn't work
was that I had desoldered two legs off one of the quad OpAmps.
Turning that thing over, side to side I started getting lost...
until undoing the mods did it become clear to me what had
happened.
That is another challenge of doing this stuff, working alone
and not in a lab or with colleagues present, we can create
problems that we don't see. Any one just looking
at it from the outside might find it instantly or with
little problem, but we just can't see it.
Like programming, writing, math, or other tasks etc.,
the challenge is looking at things from a fresh perspective.
We just have to leave it for a time, then come back and
investigate with a fresh perspective, keeping in mind it has
to be a fresh look, which is why sites like DIY and others
can be very helpful.
I'll pop the two in and see.
Not at all Dick. I put an LT1468 in there and it works great. The gain is high because the notch is better than -115dB. The input to U1 is from A3U3 notch filter output. The 1496 balance modulator would not be able track the notch at this level with out a lot of amplification. IIRC I raised the gain of U1 to get a deeper notch. And yes it does need to clip cleanly but not for long. With the LT1468 I was able to get the notch down well below -120dB. The additional gain was possible because the 1496 only need to see the fundamental. So the bandwidth of U1 doesn't need to be greater than 110kHz.
Sync you can't use the opa134 there the GBW is not high enough for a gain 200.
The amp would seriously lack bandwidth and the notch depth will suffer at high fundamental frequencies.
Sync most of what I did on the A4 board was purely experimental and didn't lead to
better performance. I did get the noise down at the IQ ( In phase / Quadrature) outputs.
There is not a lot to be gained modifying this board. If you're not clear about how it works it's better left alone.
better performance. I did get the noise down at the IQ ( In phase / Quadrature) outputs.
There is not a lot to be gained modifying this board. If you're not clear about how it works it's better left alone.
David,
Yes, I know, I've had nothing but problems with the A4 board.
However I finished demodding the board except for the LT1468.
I put back the 1K resistor, but now I'm thinking I need to put the
200 ohm back in from the mod. I do have a 12 pf cap across
R2, didn't have a 10pf.
This thing is giving me fits trying to dial in just the demod
on the board.
I'm following along your analyzer instructions from post 996
over on the long thread, namely:
HERE:
I can't set input for 1kHz at 1Vrms on TP4.
When I try it goes to never never land.
Maybe I just need the 200 ohm resistor there and try again.
I thought I've done this before but can't seem to get it right.
I tryed it Victor's too and it goes nutty. Can't think what I'm
doing wrong here.
Yes, I know, I've had nothing but problems with the A4 board.
However I finished demodding the board except for the LT1468.
I put back the 1K resistor, but now I'm thinking I need to put the
200 ohm back in from the mod. I do have a 12 pf cap across
R2, didn't have a 10pf.
This thing is giving me fits trying to dial in just the demod
on the board.
I'm following along your analyzer instructions from post 996
over on the long thread, namely:
HERE:
I can't set input for 1kHz at 1Vrms on TP4.
When I try it goes to never never land.
Maybe I just need the 200 ohm resistor there and try again.
I thought I've done this before but can't seem to get it right.
I tryed it Victor's too and it goes nutty. Can't think what I'm
doing wrong here.
I don't know if I'm reading you right here so I'll just say this.
Input to the analyzer 1kHz at 1Vrms.
We are just aligning the the notch so you leave the analyzer at a high distortion range. Not -80dB. Use the QA400 on the monitor port to watch the notch and cal the notch reasonably. Then move to the test point.
What is 'nuts'.
You have another A4 board. Put it in and try with that.
I'm assuming you didn't change the input gain.
Last edited:
TP4 of what board?
It should be U3TP4 which is the buffer output to the notch filter.
Set this point to 1Vrms @ 1kHz
U3TP3 is the output of the feedback amplifier for the notch filter.
It's just a covenant test point to measure the notch from.
I'm having to go back and look at this stuff again. It's been two years.
My instructions now will be a bit different in retrospect.
Last edited:
Put FFT on A4TP3.
I set the input for 1vrms, 1kHz.
I then read that input on A4TP4.
Confirm 1v, 1kHz.
Set the Function to distortion.
OH...maybe that is the problem, I've backed it down to
-40 dB, should have done it all the way. somewhere in
there it mentioned to -10.
The substitute board is okay, I want
to keep it for the #2.
This A4 board is the one that gives me problems.
That is why I was wondering if I should put that 200 ohm
resistor for the LT1468.
I tried the TP4 on A4 of analyzer.
Nothing on TP4 on A1. Unless I"m missing something.
I set the input for 1vrms, 1kHz.
I then read that input on A4TP4.
Confirm 1v, 1kHz.
Set the Function to distortion.
OH...maybe that is the problem, I've backed it down to
-40 dB, should have done it all the way. somewhere in
there it mentioned to -10.
The substitute board is okay, I want
to keep it for the #2.
This A4 board is the one that gives me problems.
That is why I was wondering if I should put that 200 ohm
resistor for the LT1468.
I tried the TP4 on A4 of analyzer.
Nothing on TP4 on A1. Unless I"m missing something.
Last edited:
If the board is giving you trouble then it should not be modified until repair has be carried out. You have to do these mods from a clean slate. Fix first mod later. Then you know where your coming from.
Sync your supposed to be on the A3 board for the TP3 and TP4 test points Not on the A4 board.
Then Put FFT on A4 TP1. Set input balance trim for least amount of hash and harmonics. Repeat notch calibration
monitoring the A3TP3 feedback amplifier TP.
I should have been more clear.
At the time people were following along and not confused.
Last edited:
Oh Geez,
Nuts.
Oh, yeah, that is what we told the Germans during WWII when they wanted US
troops to surrender. The C.O. said nuts, and they told that to the Germans.
I'll be back.
Nuts.
Oh, yeah, that is what we told the Germans during WWII when they wanted US
troops to surrender. The C.O. said nuts, and they told that to the Germans.
I'll be back.
Thanks, David for confirming that the 1468 is OK for U1 -- I would have suggested the LME49710 as an alternate. And, my 2cents, getting a really deep notch is a Good Thing, regardless of whether the end result is better.
Now, any reason not to use even more gain with U1? We know that performance at 100kHz isn't going to match that at 1kHz, not even by a factor of 10X, so running out of GBW at 100kHz isn't the end of the world, unless the pulses get really crappy....
Now, any reason not to use even more gain with U1? We know that performance at 100kHz isn't going to match that at 1kHz, not even by a factor of 10X, so running out of GBW at 100kHz isn't the end of the world, unless the pulses get really crappy....
I put a gain of 1000 on it and went for 90kHz.
Works very well. But a deep notch seems to take away from the tracking.
Everything is always better at 1kHz.
Works very well. But a deep notch seems to take away from the tracking.
Everything is always better at 1kHz.
Gain of 1000 works for me.
I wont be using all of BW or 90kHz
Typically 1kH, 10kHz...maybe some others.
But it's good to know it is there.
Now that I got the little on settled in....
Cheers,
I wont be using all of BW or 90kHz
Typically 1kH, 10kHz...maybe some others.
But it's good to know it is there.
Now that I got the little on settled in....
Cheers,
Gain of 1000 seems to be working.
While I was at it, I changed back
A465 back to single turn pot.
To my surprise that was another
root cause of tuning problems.
while I built it upon a 2 lead by 3 lead
perf board it was shorting.
From what I can make out the notch
is around -108 now. However distortion
only show about -65dB. Got to go back and
find out how to change internal levels to get
a better Meter reading.
Have to also figure out how to get the QA400
to read below 100 dB for noise and harmonics.
I no longer get a readout for THD or THD N.
I"ll try entering a -10dB or -20dB start point
or offset from the QA400 after a reboot and
see if that works. Anyone else run into this problem?
While I was at it, I changed back
A465 back to single turn pot.
To my surprise that was another
root cause of tuning problems.
while I built it upon a 2 lead by 3 lead
perf board it was shorting.
From what I can make out the notch
is around -108 now. However distortion
only show about -65dB. Got to go back and
find out how to change internal levels to get
a better Meter reading.
Have to also figure out how to get the QA400
to read below 100 dB for noise and harmonics.
I no longer get a readout for THD or THD N.
I"ll try entering a -10dB or -20dB start point
or offset from the QA400 after a reboot and
see if that works. Anyone else run into this problem?
Last edited:
Gain of 1000 seems to be working.
While I was at it, I changed back
A465 back to single turn pot.
To my surprise that was another
root cause of tuning problems.
while I built it upon a 2 lead by 3 lead
perf board it was shorting.
From what I can make out the notch
is around -108 now. However ditortion
only show about -65dB. Got to go back and
find out how to chnage internal levels to get
a better Meter reading.
While I was at it, I changed back
A465 back to single turn pot.
To my surprise that was another
root cause of tuning problems.
while I built it upon a 2 lead by 3 lead
perf board it was shorting.
From what I can make out the notch
is around -108 now. However ditortion
only show about -65dB. Got to go back and
find out how to chnage internal levels to get
a better Meter reading.
HP339a -2 Update
Well I was all bummed out because HP339a#2 was really
not keeping with the the problem HP339a.
I found the reason why, I left the A4 board out.
IN the problem HP339a, when adjusting it for Notch depth
I got it so low that it looks like elephant trunk with elephant ears.
Still have to investigate some of the funny business with A1U2
harmonics and the lack of 400hz filter.
Well I was all bummed out because HP339a#2 was really
not keeping with the the problem HP339a.
I found the reason why, I left the A4 board out.
IN the problem HP339a, when adjusting it for Notch depth
I got it so low that it looks like elephant trunk with elephant ears.
Still have to investigate some of the funny business with A1U2
harmonics and the lack of 400hz filter.