Do you have or can you point me to a source that does not use a transformer to “unbalance” a balanced feed?I am aware of these options, it is a simple and elegant solution (several versions are made) and while pricey, it is built to a level of quality which supports the cost.
I have spoken to both Ralph Karsten of Atma-Sphere and Josh Meredith at Cardas about these options... Ralph recommended a resister across the unused pin, and Josh asked to confirm which pin that was. I asked Ralph about my concern if it would asymmetrically load the cascode follower output stage... he said that it would in a minor way, but that it would not be a problem.
My “pre-amp” has XLR outs that conveniently connect to an XLR-in to the EXO -> XLR-out to RCA-in of the power amps. I’d like to build them inside the amps to keep things tidy
Thanks for your time and patience. Tony
Last edited:
Hi Dan, these look like quite good distortion numbers, and I salute you for taking the measurements up to 20 KHz. To me, stupendous 1 kHz THD numbers are not as meaningful (but still worthwhile).
I was confused by the the step 2 and step 3 curves. What is the difference?
What instrumentation did you use to make the distortion measurements?
Best regards,
Bob
Hi Bob,
Thanks your kind words.
For what its worth I am a fan of your book and the BAF2016 & 2019 videos were very informative. I totally agree about 20 kHz.
The lower curve is my measurement system, which is an RME-ADI2 (the black fs-r edition) AD/DA converter with Jan Didden's Autoranger Mk 2 V3 (latest firmware) running REW software.
I was running 32bit/768kHz, but distortion measurement bandwidth limited to 88.2 kHz. Signal was around 5V rms.
I have taken it out further with THD measurement and I was surprised just how good the DH220C front end is.
This design is not far above the measurement system noise floor - really impressive work Bob.
-Dan
DH-200 C1 coupling cap recommendation
I have a couple of these in front of me right now and both were put together from a kit. So, they were Hafler half soldered versions. I literally have a couple of afternoons into disassembling, messing around with wiring and soldering EVERYTHING. In any case I think I have both of these figured out and have a small list of parts in a shopping cart on Mouser. The input coupling caps C1 which are electrolytic 10mf 16v Mouser did not have in stock in the Nicichon UES. I had done some reading and I guess it is a bit of an upgrade to use a film cap in there. So, I am curious about this. I have seen some recommendations of 2.2uf and others have done 6.8uf. Are there any other modifications other than making it fit? Does anyone have a part number they have used?
I am trying just get these to a baseline of mostly stock form. However it does seem like you guys are the pros at mods. Any suggestions for easy mods that do not require several changes to do the mod? Admittedly I am a novice but learning. I have done a slew of repairs on my own, but, I don’t feel very qualified to do re-engineering.
I have a couple of these in front of me right now and both were put together from a kit. So, they were Hafler half soldered versions. I literally have a couple of afternoons into disassembling, messing around with wiring and soldering EVERYTHING. In any case I think I have both of these figured out and have a small list of parts in a shopping cart on Mouser. The input coupling caps C1 which are electrolytic 10mf 16v Mouser did not have in stock in the Nicichon UES. I had done some reading and I guess it is a bit of an upgrade to use a film cap in there. So, I am curious about this. I have seen some recommendations of 2.2uf and others have done 6.8uf. Are there any other modifications other than making it fit? Does anyone have a part number they have used?
I am trying just get these to a baseline of mostly stock form. However it does seem like you guys are the pros at mods. Any suggestions for easy mods that do not require several changes to do the mod? Admittedly I am a novice but learning. I have done a slew of repairs on my own, but, I don’t feel very qualified to do re-engineering.
Yes many Nichicon ecaps are out of stock presently. You could wait until new stock arrives. You can go up in voltage rating but check for size and fit.
Those Wimas are polyester, there are some 2.2 uf Polypropylenes that fit (and also use the bypass cap - I used FKP 0.22 for the input bypass and the output RC networks
Attachments
Last edited:
What's the purpose of adding a capacitor of ten times the impedance to the original one? Do you really think that anyone, including yourself, may notice the difference between one times and 0.909... times the total impedance?
Best regards!
Best regards!
These JFETs are available from the manufacturers: JFET, Bipolar Transistors, MOSFET, DMOS. Their website also has a list of distributors
Regards,
Ivan
Regards,
Ivan
Ah, thanks! I didn't even remember my own question in that thread regarding the output devices. Shame on me 🙄!
Anyway, those are as scarce as hen's teeth, aren't they?
Best egards!
New, 200V Lateral MOSFETs are available from Exicon as TO-3 or TO-247. Also available as dual Die versions. See: Exicon Lateral MOSFETS Distributed by Profusion, UK at: Profusion Ltd - Audio Semiconductors. They also distribute high end audio ICs from Muse as well as DSP and other audio products.
Best,
Ivan
You can also purchase matched sets of JFETs thru Rick Savas when you purchase the Cordell DH220C boards.
Is the matching between the top and bottom JFET pairs critical?
From the presentation it seemed like Bob was saying that the pairs themselves need to be well matched and the degeneration resistors at the lower pair reduce the effect of the top-bottom difference enough not to matter..
From the presentation it seemed like Bob was saying that the pairs themselves need to be well matched and the degeneration resistors at the lower pair reduce the effect of the top-bottom difference enough not to matter..
The jfet matching is checked by adjusting the IPS to 4mA and measuring/checking for VAS current for 10mA +/-2mA. You’ll find the odd ones do not meet the spec. The de-Gen R’s match the transconductances
The more closely matched they are, the better. At least the J-Fet pairs from Linear Systems are matched well, so if you can get the pairs in the ballpark with each other, life will be much easier. Don't expect a very close match.Is the matching between the top and bottom JFET pairs critical?
Rick has done a lot of very good work on these.
-Chris
Is the matching between the top and bottom JFET pairs critical?
From the presentation it seemed like Bob was saying that the pairs themselves need to be well matched and the degeneration resistors at the lower pair reduce the effect of the top-bottom difference enough not to matter..
The key is that ideally the top N-channel pair has the same mismatch amount and direction as the bottom P-channel pair. In other words, given mismatch, it is best if both pairs have mismatch of the same polarity (direction). In fact, if both pairs have significant mismatch of, say, 10 mV, that is fine if the mismatch of those two pairs is in the same direction. However, if the mismatches are in opposite directions, the net pair-pair mismatch in the circuit architecture would become 20 mV, which would be potentially serious. In such a case, just reversing the installation of one of those pairs cures the problem, with the difference in mismatches then canceling out.
Cheers,
Bob
The beauty of the design is that there is more than one method to balance it out. You can run the AFE independently. Trimming one of p-ch degen R's also does this offset correction, but in most cases reversing one of the two works just as well. That lead to installing the sot-23-6 jfets on a 6-pin DIP pcb carrier to make selection, install, service much easier. With 4 pair, as used in a stereo app, you have a few combos to work with 🙂
A lot of work 🙂 instead of using the more expensive metal can, but working with 4 pairs in the metal can, you quite possibly could run into a situation of a poor match. I have a few I can not match up, to meet spec. I suspect on the outer fringes of the variances, like high offset and a hot IDSS. I could have cooked them too 🙂 Need to measure to be sure what's up with them.
Rick
A lot of work 🙂 instead of using the more expensive metal can, but working with 4 pairs in the metal can, you quite possibly could run into a situation of a poor match. I have a few I can not match up, to meet spec. I suspect on the outer fringes of the variances, like high offset and a hot IDSS. I could have cooked them too 🙂 Need to measure to be sure what's up with them.
Rick
May I ask of which JFET's you're speaking? The only FET's I see in the original schematics are the power devices, and they aren't JFET's.
Best regards!
Best regards!
- Home
- Amplifiers
- Solid State
- Hafler DH-200/220 Mods