Much has been done and written about greening the Heathkit IG-18. But I can't find anything written about greening the HP 650 series (651-654). Is there something inherent in the design of the HP that makes it unsuitable for practical mods and improvements? It does seem to be widely available on the surplus market.
The HP 650 is circa 1956 and the manual is available at http://www.hparchive.com/Manuals/HP-650A-Manual.pdf Other manuals might be available via BAMA.
Knock yourself out, but I would think that time investment in an HP200 (I think there are 4 in the series, some standalone, some part of a telecom test set and pretty generally available) would be more worth the time.
Knock yourself out, but I would think that time investment in an HP200 (I think there are 4 in the series, some standalone, some part of a telecom test set and pretty generally available) would be more worth the time.
No, not the old 650. As I said, the 651-654. An entirely different animal. See here:
Audio & Test
They match the looks of the 330 series distortion analyzers.
Audio & Test
They match the looks of the 330 series distortion analyzers.
I've seen plenty of them all over, but that may be a regional thing.
I just downloaded the HP 651B manual and I'm shocked to see the distortion specified at 1%. Yes, one percent, I didn't miss the decimal point. Wow, that's worse than most function generators. Is that a misprint? Or are they really that bad? Hard to believe.
I just downloaded the HP 651B manual and I'm shocked to see the distortion specified at 1%. Yes, one percent, I didn't miss the decimal point. Wow, that's worse than most function generators. Is that a misprint? Or are they really that bad? Hard to believe.
The physical size makes it a bit unattractive if it's going to be just a basic test oscillator - though the attenuator, front panel, meter, controls, chassis, and variable capacitor could be salvaged and used as the basis for a home-built high purity oscillator.
I saw one at a hamfest just 2 weeks ago. It was marked $150, and the seller claimed to have no knowledge of its functional condition. That's a lot more than I wanted to spend for a nice box with some potentially useful parts.
Dale
I saw one at a hamfest just 2 weeks ago. It was marked $150, and the seller claimed to have no knowledge of its functional condition. That's a lot more than I wanted to spend for a nice box with some potentially useful parts.
Dale
I can't see a reason in the world that it would be that bad. The old 200 series with tubes were used by McIntosh for their amplifier clinics back in the day. They claimed to have modified it slightly for lower distortion, IMO this one should be way better than that- unless they lost something by increasing the maximum frequency capability. I'd start by looking at the amplitude stabilization scheme (haven't looked a the circuit). Does it use bulbs? They're really very good- add more bulbs for time constant. Look at what Jim Williams had to do to get rid of them and still get good performance in his oscillator designs. Also look at the oscillator circuit Bob Cordell uses in his distortion analyzer for ideas. Personally I keep a 204 around and rarely use my 200 anymore- it's a pristine rack mount version and I mostly keep it as a display piece.
I'm hoping someone will chime in who is familiar with the 650 series. I just had a casual look at the schematics. On paper it's an order of magnitude more sophisticated than the Heathkit. Example: Heathkit power supply single rail, one transistor. Very basic. HP, 2-rail, 7-8 transistors depending on model, plus a couple opamps in one model. Fold-back current limiting separately on each rail. Output ripple specified at 1.3-1.7 mV.
It just doesn't make sense that the HP would be an order of magnitude worse distortion than the Heath.
It just doesn't make sense that the HP would be an order of magnitude worse distortion than the Heath.
These oscillators cost from $20 to $80, perhaps less at hamfests. As I stated, the mechanicals are unbeatable.
One thing I aim to do is use the Linear Tech "true rms" detector LTC1966 with a VCA in one of the HP204's, adjusting the time constant of the LTC1966 with the range switch. One might also consider an ultimate POOGE by implementing the Krohn Hite 4400 VCA. I wrote an article for AX describing the useage of these RMS detectors, and they are very useful indeed.
http://cds.linear.com/docs/Application Note/an106f.pdf
One thing I aim to do is use the Linear Tech "true rms" detector LTC1966 with a VCA in one of the HP204's, adjusting the time constant of the LTC1966 with the range switch. One might also consider an ultimate POOGE by implementing the Krohn Hite 4400 VCA. I wrote an article for AX describing the useage of these RMS detectors, and they are very useful indeed.
http://cds.linear.com/docs/Application Note/an106f.pdf
HP651A
I acquired a used HP651A in very good condition about 15 years ago with all the manuals and it works perfectly. Rock solid output 10 Hz - 10 MHz.
I have measured THD in the audio range and its under 0.1%. I'd have to dig back through some old notes to quote the HD spectrum.
0.1% is OK for general purpose use, but its too high for testing HD on serious audio amplifiers, and 10 Hz is not really low enough.
IIRC, it uses a transistor based wein bridge followed by a transistor base line driver stage.
I acquired a used HP651A in very good condition about 15 years ago with all the manuals and it works perfectly. Rock solid output 10 Hz - 10 MHz.
I have measured THD in the audio range and its under 0.1%. I'd have to dig back through some old notes to quote the HD spectrum.
0.1% is OK for general purpose use, but its too high for testing HD on serious audio amplifiers, and 10 Hz is not really low enough.
IIRC, it uses a transistor based wein bridge followed by a transistor base line driver stage.
Thanks Glenn, that answers one part of the puzzle. So the actual distortion is in fact 0.1%, not the advertised 1%. Yes that does still seem a bit lame considering the unmodified Heathkit is about the same.
Still wondering why nobody's taken a shot at improving it. Looks like a better starting platform than the Heathkit.
Still wondering why nobody's taken a shot at improving it. Looks like a better starting platform than the Heathkit.
HP651A
Looking at page 1 of http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1965-06.pdf , the 1% distortion is up near 10 MHz.
Is it worth me scanning & posting the schematic?
Looking at page 1 of http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1965-06.pdf , the 1% distortion is up near 10 MHz.
Is it worth me scanning & posting the schematic?
The block diagram on pg 4 shows a few things to think about:
- It's a capacitor-tuned Wien-bridge circuit, like the HP200 et seq, which seems to be the most common approach (perhaps the only practical topology?) for getting frequency spans from audio into the MHz range.
(Hmmm . . . maybe not . . . I think some of the early General Radio and Boonton generators did it by heterodyning a tunable RF oscillator against a fixed oscillator.) - The passive elements have 2:1 value ratios rather than equal values. I don't recall how this affects harmonic suppression by the Wien network, but I think this increases the gain required from the amplifier. By using an equal-value Wien network the amplifier gain - and, presumably, distortion - can be reduced.
- Level control uses some kind of a diode network. Unless the signal amplitude is VERY low in the ALC circuit, I can't imagine that this has better distortion performance than other alternatives.
- The output amplifier is capable of driving 3.16V (RMS) into a 50 ohm termination. The 600 ohm output impedance is achieved by adding series resistance. It may be possible to improve the distortion into 600 ohm loads by sacrificing some of the 50 ohm capability.
Well, that's on a tangent, but several reasons.
1. The 239A is not cheap when available, pretty much same price as a 339A in my experience.
2. I enjoy tinkering (after all it's a hobby) and the 239 and others like it are all you need, no need to modify
3. I already own a Heathkit, which I'm about to mod. There is a large body of info from those who have spent a lot of time and effort improving it. If the same kind of info were available on the HP 650 series, I wouldn't mind ditching the Heath and buying an HP to mod. They're basically $50 and easily available. On the other hand, I don't have the time to endlessly trial-and-error mods to the HP on my own.
1. The 239A is not cheap when available, pretty much same price as a 339A in my experience.
2. I enjoy tinkering (after all it's a hobby) and the 239 and others like it are all you need, no need to modify
3. I already own a Heathkit, which I'm about to mod. There is a large body of info from those who have spent a lot of time and effort improving it. If the same kind of info were available on the HP 650 series, I wouldn't mind ditching the Heath and buying an HP to mod. They're basically $50 and easily available. On the other hand, I don't have the time to endlessly trial-and-error mods to the HP on my own.
I agree, so it would be interesting if there was a simple mod for the HP 650 for a significant improvement, then one could build a 239 into that chassis as an all out alternative mod.
Here is a scratch build of the HP239A into a Heath:
http://www.tronola.com/html/ig-18_mods.html
It includes a source for the Harris OP amp.
There is of course also Bob's oscillator, does anyone know which has lower distortion at 20 kHz?
Here is a scratch build of the HP239A into a Heath:
http://www.tronola.com/html/ig-18_mods.html
It includes a source for the Harris OP amp.
There is of course also Bob's oscillator, does anyone know which has lower distortion at 20 kHz?
Last edited:
Dick Moore (richiem user ID here on diyAudio) discussses some of the better oscillators out there on this page, with brief mention of the HP651 series:
About RC oscillators
Most of the early HP oscillators with the large tuning indicators are capacitance tuned Wien Bridge oscillators, the later ones are balanced Wien Bridge types. The tuning capacitors are typically on the order of 600 pF and very large resistors (10 to 50 Meg ohm) are required for the lower frequency ranges. I'd think that the non-linearity of stray impedances would limit the performance of these units. I'm surprised that HP retained the tuning cap for so long.
I happen to have an HP 202C which is a tube, balanced Wien Bridge type - built like a tank.
Jim Williams notes in AN-43 (Figs 47 and 48 which are inadvertently reversed) that distortion in Wien Bridge type oscillators can be greatly reduced by nulling the common mode voltage at the input and he shows how to do it with an OP amp.
About RC oscillators
Most of the early HP oscillators with the large tuning indicators are capacitance tuned Wien Bridge oscillators, the later ones are balanced Wien Bridge types. The tuning capacitors are typically on the order of 600 pF and very large resistors (10 to 50 Meg ohm) are required for the lower frequency ranges. I'd think that the non-linearity of stray impedances would limit the performance of these units. I'm surprised that HP retained the tuning cap for so long.
I happen to have an HP 202C which is a tube, balanced Wien Bridge type - built like a tank.
Jim Williams notes in AN-43 (Figs 47 and 48 which are inadvertently reversed) that distortion in Wien Bridge type oscillators can be greatly reduced by nulling the common mode voltage at the input and he shows how to do it with an OP amp.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.