Hi y'all,
I fat fingered the title, it should only say HP 339A, w/o the 8 or more like the following:
Or Does anyone have any in their stash that they could part with that would be an upgrade to the stock opto's that HP used?
If so, would you be kind enough to PM me?
I'm not sure how to measure and check the opto couplers and find those with low distortion for use in the HP339A.
Cheers
I fat fingered the title, it should only say HP 339A, w/o the 8 or more like the following:
HP339A Optocouplers and trouble shooting.
If tried and it doesn't work well enough, then does anybody have any suggested replacements? Or Does anyone have any in their stash that they could part with that would be an upgrade to the stock opto's that HP used?
If so, would you be kind enough to PM me?
I'm not sure how to measure and check the opto couplers and find those with low distortion for use in the HP339A.
Cheers
Hi,
When I first stumbled into this forum, I assumed you guys were trying to repair test equipment. But now I understand you’re trying convert derelicts into Hot Rods. BRAVO!
Re testing optocouplers: there’s useful material on distortion in this link: PerkinElmer Optoelectronics - datasheet pdf
If I understand correctly, you have a working HP339A. If so, you can use it to test candidate couplers’ distortion. In the following, keep in mind that the generator output impedance is 600 ohms.
Say you were evaluating parts for the A4E2 coupler depicted in Fig. 8-6 of the service manual. Use the oscillator output to drive a 2.7K resistor (i.e. ~ 3.32k - 600 ohms) in series with the photocell. You’ll be measuring the distortion at the junction of these two parts. Use a variable bench supply, filtered, to drive the LED--- e.g. two 220 ohm resistors in series with a 100 uF or larger cap between them to ground. The supply must be clean to avoid introducing distortion into the measurement. Start with the LED supply at 0V and set oscillator output at 0.2V peak. Then raise supply voltage until the signal on the photocell drops in half and measure the distortion. The photocell resistance is now about 3.3K and the coupler-generated distortion can actually be thought of being twice the displayed analyzer reading because the distorting errors are being cut in half by the equal value resistances. The residual distortion of the analyzer will less than this pessimistic measurement might suggest because the coupler introduces a small tweak on other resistors. But this should be a valid technique for A vs. B comparisons.
Interesting experiments would be to vary oscillator amplitude and LED current. My intuition is that worst case distortion will occur roughly when LED drive reduces source amplitude in half. An ability to handle larger signal amplitude at comparable distortion would probably constitute an improved coupler.
All this said, I’m not optimistic that you’ll be able to find an optocoupler that leads to lower overall analyzer residual. I suspect improvement will have to come from improved circuitry rather than better parts. Which leads me to a question for you and other community members.
I’ll pose the same question in a couple of different ways.
If the HP339A analyzer were driven by a “perfect” oscillator with 0 distortion, does anyone have a feel for what the analyzer residual measurement floor might be?
Assume the HP339A oscillator had already been improved with the techniques described by your colleagues. If it was then possible to manage an additional reduction of oscillator distortion by about 6dB, would that be worth pursuing?
I confess to getting lost in the weeds of the forum. Is there any summary of best mods to the HP339A oscillator and the resulting overall performance?
When I first stumbled into this forum, I assumed you guys were trying to repair test equipment. But now I understand you’re trying convert derelicts into Hot Rods. BRAVO!
Re testing optocouplers: there’s useful material on distortion in this link: PerkinElmer Optoelectronics - datasheet pdf
If I understand correctly, you have a working HP339A. If so, you can use it to test candidate couplers’ distortion. In the following, keep in mind that the generator output impedance is 600 ohms.
Say you were evaluating parts for the A4E2 coupler depicted in Fig. 8-6 of the service manual. Use the oscillator output to drive a 2.7K resistor (i.e. ~ 3.32k - 600 ohms) in series with the photocell. You’ll be measuring the distortion at the junction of these two parts. Use a variable bench supply, filtered, to drive the LED--- e.g. two 220 ohm resistors in series with a 100 uF or larger cap between them to ground. The supply must be clean to avoid introducing distortion into the measurement. Start with the LED supply at 0V and set oscillator output at 0.2V peak. Then raise supply voltage until the signal on the photocell drops in half and measure the distortion. The photocell resistance is now about 3.3K and the coupler-generated distortion can actually be thought of being twice the displayed analyzer reading because the distorting errors are being cut in half by the equal value resistances. The residual distortion of the analyzer will less than this pessimistic measurement might suggest because the coupler introduces a small tweak on other resistors. But this should be a valid technique for A vs. B comparisons.
Interesting experiments would be to vary oscillator amplitude and LED current. My intuition is that worst case distortion will occur roughly when LED drive reduces source amplitude in half. An ability to handle larger signal amplitude at comparable distortion would probably constitute an improved coupler.
All this said, I’m not optimistic that you’ll be able to find an optocoupler that leads to lower overall analyzer residual. I suspect improvement will have to come from improved circuitry rather than better parts. Which leads me to a question for you and other community members.
I’ll pose the same question in a couple of different ways.
If the HP339A analyzer were driven by a “perfect” oscillator with 0 distortion, does anyone have a feel for what the analyzer residual measurement floor might be?
Assume the HP339A oscillator had already been improved with the techniques described by your colleagues. If it was then possible to manage an additional reduction of oscillator distortion by about 6dB, would that be worth pursuing?
I confess to getting lost in the weeds of the forum. Is there any summary of best mods to the HP339A oscillator and the resulting overall performance?
A summery was identified as post #1184 in the long distortion thread. That is the link from the blog and wiki.
This is RN Marsh's HP339A distortion measurement of the modded oscillator and other components; work based on the late RichEEM, Davada, 1Audio, RNMarsh, and others. The Blog link is here: LINKTOHP339a PIC
Something around -132dB or the like. Real link to the blog mod: LINK
Here is the fig 8-6 for anyone following along.
This is RN Marsh's HP339A distortion measurement of the modded oscillator and other components; work based on the late RichEEM, Davada, 1Audio, RNMarsh, and others. The Blog link is here: LINKTOHP339a PIC
Something around -132dB or the like. Real link to the blog mod: LINK
Here is the fig 8-6 for anyone following along.
Attachments
Thanks, Sync.
I'm going to follow up with a marked-up schematic detailing mods as I understand. Thinking that might be easiest way to review and clarify. Probably won't happen until tonight.
Do you recommend moving this exchange to the huge thread on low distortion oscillators since I'm going to focus possible HP339A oscillator improvement?
Thanks again.
I'm going to follow up with a marked-up schematic detailing mods as I understand. Thinking that might be easiest way to review and clarify. Probably won't happen until tonight.
Do you recommend moving this exchange to the huge thread on low distortion oscillators since I'm going to focus possible HP339A oscillator improvement?
Thanks again.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.