Split from the Blowtorch thread
Hi John,
Thanks, but the shipping would kill me.
At this point, is there any reason to hide your actual schematics? It's an old design that depends on discontinued devices and careful hand construction. The PCB details are enough to stop a clone.
The reason I ask is that you could refer to specific design examples as you pointed out your design choices, or what you consider to be important concepts. It would be somewhat easier to keep things on track and clear to other people.
Considering the difficulties a normal person would have to begin production on anything, there really is not a lot to worry about there either. I do not want to make one of these, but learning from this would be very useful to many.
Just a thought John.
To create better audio devices, I think better equipment is in order these days. Sure, I can see all kinds of stuff using the gear I have now, but either a rebuild of my current stuff, or new equipment would be required to look at any cutting edge analog stuff that is current.
My problem is that I don't know where to begin to assess what to do with the existing circuitry. The guys at HP/Agilent know a heck of a lot more than I do about wide band, low noise circuitry.
Hi Scott, (or anyone else),
There is no reason why you would want to do this, but I'll ask if you have any ideas on how to improve the signal amps in the following HP equipment:
339A
3580A
3581A
3585A
or others that many people may have.
This would allow many more DIYers to get better measurements from equipment they already have. I would love to extend the service life and S/N ratio of these instruments. I imagine that the signal amps may be similar between some of these. I understand that John has improved his Sound Technology stuff, but that was op amp based, wasn't it?
Yes, a different thread. I just wanted to get Scott's attention. I asked Scott because I feel he has the knowledge and skill to pull something like this off. I imagine some other people would also have the skills required.
-Chris
Edit : Spelling
Hi John,
Thanks, but the shipping would kill me.
At this point, is there any reason to hide your actual schematics? It's an old design that depends on discontinued devices and careful hand construction. The PCB details are enough to stop a clone.
The reason I ask is that you could refer to specific design examples as you pointed out your design choices, or what you consider to be important concepts. It would be somewhat easier to keep things on track and clear to other people.
Considering the difficulties a normal person would have to begin production on anything, there really is not a lot to worry about there either. I do not want to make one of these, but learning from this would be very useful to many.
Just a thought John.
To create better audio devices, I think better equipment is in order these days. Sure, I can see all kinds of stuff using the gear I have now, but either a rebuild of my current stuff, or new equipment would be required to look at any cutting edge analog stuff that is current.
My problem is that I don't know where to begin to assess what to do with the existing circuitry. The guys at HP/Agilent know a heck of a lot more than I do about wide band, low noise circuitry.
Hi Scott, (or anyone else),
There is no reason why you would want to do this, but I'll ask if you have any ideas on how to improve the signal amps in the following HP equipment:
339A
3580A
3581A
3585A
or others that many people may have.
This would allow many more DIYers to get better measurements from equipment they already have. I would love to extend the service life and S/N ratio of these instruments. I imagine that the signal amps may be similar between some of these. I understand that John has improved his Sound Technology stuff, but that was op amp based, wasn't it?
Yes, a different thread. I just wanted to get Scott's attention. I asked Scott because I feel he has the knowledge and skill to pull something like this off. I imagine some other people would also have the skills required.
-Chris
Edit : Spelling
Hi John,
Thank you. I was afraid of that. Still, I would like to keep an open mind about these things.
I wouldn't go ahead and work on one without a clear idea of what I needed to do. Even things like temperature compensation have been well thought out. I have a ton of respect for the engineers who designed this stuff.
There may be some small gains in working over some power supplies.
-Chris
Thank you. I was afraid of that. Still, I would like to keep an open mind about these things.
I wouldn't go ahead and work on one without a clear idea of what I needed to do. Even things like temperature compensation have been well thought out. I have a ton of respect for the engineers who designed this stuff.
There may be some small gains in working over some power supplies.
-Chris
Let me talk about professional test equipment a little. HP, TEK, and many others make excellent test equipment and they do not go cheap on the parts. If they could find a part when the unit was first made or a different circuit that would give them better performance, they would have done so. Not true, necessarily in consumer audio or pro audio, and that is why it is easier to modify and improve.
To think that amateurs, or even professionals, without test equipment to measure noise, distortion, and voltage breakdown of individual devices, can easily and safely improve most professional test equipment is the height of folly, and if you were a soldier, the same mentality would quickly get you killed. That is why they have old sergeants to keep them in line. The biggest problem is usually self noise which can be higher than optimum. However, a X 100 gain, IC based low noise gain stage on batteries for low and difficult measurements, can overcome the vast majority of compromises in available test equipment. This is why I think that modifying a wave analyzer, or scope is a waste of time.
To think that amateurs, or even professionals, without test equipment to measure noise, distortion, and voltage breakdown of individual devices, can easily and safely improve most professional test equipment is the height of folly, and if you were a soldier, the same mentality would quickly get you killed. That is why they have old sergeants to keep them in line. The biggest problem is usually self noise which can be higher than optimum. However, a X 100 gain, IC based low noise gain stage on batteries for low and difficult measurements, can overcome the vast majority of compromises in available test equipment. This is why I think that modifying a wave analyzer, or scope is a waste of time.
Joshua_G said:
Joshua,
For a good number of reasons, this is way over your head. Leave it as is, or ask somebody which is skilled in HP service.
You can download the service manual from the Agilent site and do some lecture about.
Joshua_G said:
Not all of those opamps have the datasheets readily available, so finding better options is not easy. Most of them are in the old TO-99 round can, almost impossible to find today, so you need some sort of adapter (with added inductances and capacitances, depending on the chosen solution).
E.g. you can certainly find a replacement for LM348 (quad 741) but not so easy for the "1826-0487" opamp.
Perhaps some kind sould that already did and evaluated the upgrades will help you, otherwise this is not something you should expect for free. The costs in work and parts could be a significant chunk of what you originally payed for the 339A equipment. Then you have to recalibrate the instrument, another significant task, running the total costs up high.
So, one to another, you either have the significant skills and the time to DIY, or forget about. It's not a plug-and-go job.
For the HP3581 I use a battery operated JFET preamp -- as JC recommends. I just use an LSK389 followed by an LT1028 (sorry Scott). It's helpful to bandwidth limit the preamp to 50kHz. the design is similar to Denis Colin's in AX, but without "A" weighted filtering.
One of these days I will replace the NiCad battery pack on the device.
I have connected the HP 3581 to a PC with a Dataq acquisition module.
One of these days I will replace the NiCad battery pack on the device.
I have connected the HP 3581 to a PC with a Dataq acquisition module.
Hi John (syn08),
I agree with you on those points.
Mind you, given the age of these circuits, its hard to imagine we can't do any better today. Mind you, that must be done carefully or we'll end up with lower performance.
I think there is a cross reference for the HP parts back to the original types. In fact, that information may be in the manual in the parts section.
-Chris
I agree with you on those points.
Mind you, given the age of these circuits, its hard to imagine we can't do any better today. Mind you, that must be done carefully or we'll end up with lower performance.
I think there is a cross reference for the HP parts back to the original types. In fact, that information may be in the manual in the parts section.
-Chris
I have been hot rodding distortion analyzers for some years now and you can improve them some but not necessarily a lot. And there are a lot of locations to look at. The Boonton 1120 that I have a lot of experience with has several performance limits that are really constrained in the design. Place that will be problems with other units. For example a commercial analyzer must have a very large input range, usually from 1 mV or less to 300 V. And it must be protected. That requirement really limits the noise floor of the circuitry, or makes it very fragile.
The lowest noise or distortion opamp won't necessarily drop in and work. Input bias or current noise may be a problem because of the high impedance divider circuitry.
There will be filter circuitry that can impact the performance. On the Boonton the high pass filter limits the distortion floor and increases it from .0007 to .0018 or so. Better opamps make no difference. Its too complex to worry about so I looked elsewhere.
Not every opamp in the notch circuit contributes to residual so replacing them all is a waste of money.
Also the lowest distortion may not come from the "best sounding" opamp.
Improving scopes is not going to get very far. The CRT is a limit, as is the vertical amps etc. A TEK 7633 like John Curl has is a good balance in value and with the plug in options very flexible. Its possible to get up to 1 GHz real time in the Tek 7104 and 20 GHz+ with sampling plugins if you need to . . . And the 7A22 has enough sensitivity to see almost anything. Getting a Tek 7K scope and appropriate plugins is a better value that trying to improve a scope that doesn't do what you need.
The lowest noise or distortion opamp won't necessarily drop in and work. Input bias or current noise may be a problem because of the high impedance divider circuitry.
There will be filter circuitry that can impact the performance. On the Boonton the high pass filter limits the distortion floor and increases it from .0007 to .0018 or so. Better opamps make no difference. Its too complex to worry about so I looked elsewhere.
Not every opamp in the notch circuit contributes to residual so replacing them all is a waste of money.
Also the lowest distortion may not come from the "best sounding" opamp.
Improving scopes is not going to get very far. The CRT is a limit, as is the vertical amps etc. A TEK 7633 like John Curl has is a good balance in value and with the plug in options very flexible. Its possible to get up to 1 GHz real time in the Tek 7104 and 20 GHz+ with sampling plugins if you need to . . . And the 7A22 has enough sensitivity to see almost anything. Getting a Tek 7K scope and appropriate plugins is a better value that trying to improve a scope that doesn't do what you need.
Joshua, I don't have the 339 schematics directly available to me, but I have used them and they are comparable to the ST analyzers, that I have modified. Two areas come to mind, the oscillator IC's and the input IC's. Adaptors can be purchase, but I now use adapted minidip sockets with added wires. Works OK. You can lower the residual oscillator distortion, and the input noise by as much as 10 dB. It works! However, this is the only HP product that I would recommend modification. It is generally better just to make a little perhaps gain switchable quality op amp gain stage for very low inputs. Signal averaging can be a wonderful fix for distortion resolution, even 10X is good, 100 times is wonderful.
What is the frequency response we need into an osciloscope?
Also, if we need a very high frequency responsive instrument, what is the reason for that need, when audio goes to 20K and interesting harmonics does not goes above 1 Megahertz.
The reason of this question is that i have a 20 Megahertz one, and i would like to know their limiting factors... also if only the frequency response is the an important limiting factor.
regards,
Carlos
Also, if we need a very high frequency responsive instrument, what is the reason for that need, when audio goes to 20K and interesting harmonics does not goes above 1 Megahertz.
The reason of this question is that i have a 20 Megahertz one, and i would like to know their limiting factors... also if only the frequency response is the an important limiting factor.
regards,
Carlos
An externally hosted image should be here but it was not working when we last tested it.
1audio said:
The output stage of the Boonton, at least in the later series of analyzers, is very low noise, very low THD% -- I looked at it with my AP and it was about at the floor of the latter. The whole source chain is very good.
I concur with the opamps -- I guess mine has Signetics -- there is no point in switching them out.
The Boonton 1120 benefits from switching out the filter caps on the power supply -- there is a fellow who worked there (they are in NJ) who told me that the old caps were prone to getting over-worked. I don't think that there is anyone at their facility in Parsippany who is old enough to remember when the 1120 was in production, but they will calibrate it for you. As you can guess, I had mine calibrated over there (it's about 10 mi from where I live).
An analyzer gives you somewhat helpful data -- but if you have a spectrum analyzer to look at the distortion residual you'll be in much better shape. It doesn't have to be an expensive SA.
One thing which will really help with older DVM's is to use a true RMS converter. Most hand-helds get useless above a few kHz, even looking at sine waves. There have been a couple articles on deploying the Linear Tech converters -- the application note from the company is even more helpful!
Wrt the 7a22n -- I use a 5a22n in a Tek 5223. this is a wonderful plug in. I had a 7704 and a 7603 but they took up too much space.
I've been troubleshooting and designing audio stuff for decades with nothing more than my trusty Tek 545B, high gain differential plug-in and various meters. In terms of performance, my home built Cordell THD analyzer can't be beat by any HP stuff. For anything over 30MHz I use other techniques, but it's rarely necessary with audio gear other than tuners, and those are quickly going the way of the dodo as Lo Definition radio takes over. Any good piece of test equipment is a *system* so random spot improvements don't buy much of anything. The people who designed Tek, HP/Agilent, GR and various other top-line test equipment were very good and were working in an economic time when they had far more freedom to fine tune the designs and obtain custom components. The design process is different today. We have extremely high performance ICs that integrate lots of functions, we do little component level design and have to spend a lot of time on software that might not have been part of an older instrument at all. In either case it's hard to go in and perform some upgrade that makes much difference. Finally, a clean unmodified instrument is worth far more than one with what I affectionately refer to as "technician blight", for which the garden store has no chemical remedy.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.