I was going to build one of these when they first came out, I really would like a Pass Labs HP1-A but its out of my budget.
I sent a note to Pass asking them if they would bulid a HPA-1 with out the preamp out at a lower cost and they said, the preamp is not a big part of the amp or cost? and recommended me to lk at the whammy.
Regardless of this I have built many DIY amps, and have many retail amps as well.
I did alot of pp amp testing with a Burson FUN amp…Class A 2 watts with several op amps. I did the Burson V6 and V7, Sparkos, and several others…it was a very Clean. crisp sounding amp. And IMO a great deal for the price.
PCB will be here today, the parts were pretty much all there in stock at Mouser except for the 1N4004 diodes, but they had several subsitutions and the transformer I got at digi key the 70065K with 22v secondarys.
The case will be availble from DIY in approx 1 months…I will wait for that, I think.
Ugh I’m looking for the changes to lower the voltage. I know they’re berried somewhere in the 194 page build thread on DIYaudio but don’t remember where exactly.
Indeed I use 22V transformer and LED configuration.
My PSU outputs +/-16.7V, and with the quite high consumption from the bursons, I got +/-15.4V at Op Amp socket.
From what I understand, the PSU is not critical thanks to the highly efficient CRCRC circuit. So I would not sweat it too much.
Note that I still added a filtering power socket because why not.
Indeed going naked regulator config is the best if you really want to try many Op Amps, but honestly there are plenty supporting more than +/-15V out there.
Here is the list of mods I did:
Removed input caps (replaced by wire). Pay attention that your source must then have little to no DC. I have less than 0.3mV of DC on my outputs with my preferred Op Amps (Op Amps can also introduce their own DC, but it should remain very low as soon as you keep C26/27 and do not short them.
Bypassed all capacitors on signal path by some film capacitors, typically Wima MKP10 1uF. (C12/17/26)
Decreased output resistors to increase the bias from 60mA to 80mA. R16/22//29/32 changed from 10ohms to 6.88ohms (just soldered a 22ohms on top of the 10ohms). This is the result of quick experiments, works great, and can easily be reverted. Note that I went for a bit taller heatsinks (8cm instead of 5cm tall if I recall) already from the start since I used the Galaxy U2 chassis.
Doubled the value of C26 to decrease even further down the low frequency extension. This was most probably not needed and might create more leakage current issue than anything else (with sensitive Op Amps), but in my case it still performs great.
Beefed up a LOT the decoupling around the Op Amp socket, with extra Panasonic FC and FR, and some film cap with short leads, 0.33uF MKP10 between V+/V- and 0.1uF MKP4 between V+/GND and V-/GND. I felt this was a nice to have since I had to compare some Op Amps with higher consumption.
Beefed up decoupling capacitors close to output stage with mixtures of electrolytic and film caps on C9/28/6/11 and even C8/C10 in an unverified attempt to have a more " stable" LED reference (I have read somewhere that the small cap was needed to help smooth the irregular Vdrop across the LED). Cannot hurt
increased R20/R21, decreased C18/19. This results in a bit less heat on the regulators and better CRCRC filtering. In addition it also taxes less the transformer with less inrush current, “softer” recharge if you will.
I used some bitumen and blue tack that was laying around to dampen the parts of the PCB / chassis that ringed when knocking on them.
Beefed up the heatsinks on the warmest MOSFETS/Regs in an attempt to have them all working at close temperatures. Note: I did not dive to look for P and N sweet spot for Vds/Ids/Temperature, or I would be spiting hairs there (or was I already? )
Electromagnetic shielding on transformer and high impedance/ sensitive parts.
Among those mods, the two biggest improvements were obtained from Bias output increase and removing/bypassed signal path capacitors. (Yes I had the patience to do them step by step and listen all day at work).
Great, 6L6 repsonded to my query…and said using a 22v sec xfmr was fine with the not using D5/D6 led diodes and jumpering R9/13…viola 15 volts…(ie config 2, the naked route).
I have both xfmrs coming…15 and 22…so i wil decide which way to go…maybe i will build a second one depending how i like the first one.
Thank YOU for your mod list…i like these and like u will listen with the amp as is and then try these out one at a time etc…
Look forward to sharing with you mu build!
Alex
Oh did u actually measure the bias or just do the calculation and change resistor value…?
Also what change or improvement was there?
Womdering if this would affect the volume level at the same phsical positions etc?
I still have the classic in the amp.
It can be a little warm sounding, but I generally prefer it over the other opamps I’ve tried.
I’ll probably pick up a V6 vivid for the amp at some point .
Well its completed up and running with an OPA2134, drives both my HEDDS and Focal Clears very well. Actually its pretty dang nice…its a keeper here for sure.
So far I have tried several op amps in the Whammy and am happy to report they all work very well, no
issues with oscillations etc…
The amp drives my Focal Clears, Meze Emperyeans, HD 600’s and HEDD headphones all very nicely.
So far I have tried, LM 833, AD746, NJR2068, NJR4556, OPA2107, OPA2134, and OPA1612 (SOIC-8).
All except the 1612 are the old 8 pin dips…
There are a few OP Amps that arent made in 8 pin dip package any more but u can get the single version and wire up or solder to a pcb to get the same result.
Also the newer op amps designed especially for headphone amps (TI) only come in the smaller SOIC - 8 and smaller packages…making it harder for DIY folks to use…
They all work well…the differences are in the “splitting hairs” region…if u did dare to pick one or two the OP2134, or
AD746.
Observation with other amps: the biggest one is power available, the Whammy is 400 mw or so maybe up to a watt at 32 ohms.
That said its a beautiful sounding amp…very clear etc. As good as any other amp like the Schiit series, more clear than the Asgard 3 IMO. more so than the Lyr 3, similar to the Jotenheim.
It has enough power to drive the Focal Clears easily the Meze Emperyeans and even the more hard to drive HEDD’s…not an issue at all.
Most of the time is spent with the OP2134 and AD746, So this might tell u something.
After several weeks now with the Whammy and my headphones, I decided to lower the gain from the default 5.75 x to closer to unity gain.
Replacing the input resistors for the op amp from 1K to 10 K lowers the gain to 1.5, close to unity.
Why i did this was because of two things:
Using the lowest gain is best for the least amount of distortion.
Get the volume pot out of the low portion of the pots mechanical position.
WIth the default gain, the volume pot is at 8 -9 oclock, now its 10 to 12 oclock for decent listening.
Your out of the pots lower area and away from any low level mis-tracking issues as well…and the amp works very well in this new position.
Some electrical engineers mention the amp “works” better in the middle to upper levels of a pots travel…
Anyhow its done and all is well…and the pairing with the Empys now “seems” superlative!!
)
