I revisited the ACA amplifier, as not being confident, I lucky to find a missing input resistor that once placed, tame the high sensivity issue. Now volume steps are on a par within HP and Line.
Back to the DCG3, I see in this desing M3 transistor is a DN2540, but I wonder if is posible instead place an Ixys like 1N100, that used to achieve better sound in SSHV2 reg.
M1 and M2 are mosfet IRF9610 which still perform well after years, never worn out transistors? may have those also an improuvment? I am wondering if some NOS Jfet are worth to look for and try.
Back to the DCG3, I see in this desing M3 transistor is a DN2540, but I wonder if is posible instead place an Ixys like 1N100, that used to achieve better sound in SSHV2 reg.
M1 and M2 are mosfet IRF9610 which still perform well after years, never worn out transistors? may have those also an improuvment? I am wondering if some NOS Jfet are worth to look for and try.
Missing R11? So your ACA amps were running open loop max gain. That explains the abnormal sensitivity. Problem solved.
Replacing power semiconductor types may destabilize and/or wrongly bias the DCG3. Can also worsen THD. Unless done under lab conditions properly documented as adequate.
To change the sound,
the uPA68 can be swapped for a pair of matched quad 2SK117GR of ample IDSS.
The bc560 quads can be swapped for BC327-40, maybe a little more relaxed sound. Not sure about the others.
There are a lot of ways of doing a 26, isn't a kit that you only have to solder components.....
Hi, just finished right now the shootout, bc560 are not tested, guess Ok as are coming from parts kit. BC327-40 lukily I have plenty, picked a paired strong idss ( 44,4 and 44,2).
In advance I placed three pin sockets for a convenient quick removal, also I add to the bias trim an small adhesive tape, to help set correct bias values for each transistor pair in a quick swap.
Music is vinyl record 'La Spagna' by Bis label, is ancient music with a great variety of instruments, well played and well recorded. Good for evaluation.
First listen is bc327-40, I could'nt find any weakness, the flow is so natural I let the entire side on play.
After swap to bc560 the listen not same, soloist instruments much less rich timbrically, it is a glare I would not say metallic or agressive, not at this extreme, just a dusty or grey glare, that makes boring the listening to this demanding music. Other parts more complex, armonies between instruments, sound even worse without authority, much weaker sounding in a word. That prove evident after a quick third listen, back to the bc-327-40.
In advance I placed three pin sockets for a convenient quick removal, also I add to the bias trim an small adhesive tape, to help set correct bias values for each transistor pair in a quick swap.
Music is vinyl record 'La Spagna' by Bis label, is ancient music with a great variety of instruments, well played and well recorded. Good for evaluation.
First listen is bc327-40, I could'nt find any weakness, the flow is so natural I let the entire side on play.
After swap to bc560 the listen not same, soloist instruments much less rich timbrically, it is a glare I would not say metallic or agressive, not at this extreme, just a dusty or grey glare, that makes boring the listening to this demanding music. Other parts more complex, armonies between instruments, sound even worse without authority, much weaker sounding in a word. That prove evident after a quick third listen, back to the bc-327-40.
Attachments
Maybe you mean hfe (beta) 444 & 442 for the BC327-40 pair? Because bipolar transistors its impossible to have Idss. That's for JFET & DMOS.
Technically the BC560-C gives more open loop gain when in this current mirror which translates to more feedback and its also a faster type. When the BC327-40 gives some lower open loop and its slower but has significantly less inherent noise. Especially in the 1/F area of low mids and lows.
Technically the BC560-C gives more open loop gain when in this current mirror which translates to more feedback and its also a faster type. When the BC327-40 gives some lower open loop and its slower but has significantly less inherent noise. Especially in the 1/F area of low mids and lows.
Continuing with the build. Received the enclosure. The quality is very good, and there are some convenient holes for the source selector for leds which were not visible in the images when placing the order. The color is natural aluminium for the faceplate, although it looked more white in the pictures, still nice though. The knobs, source selection and volume control, are surprisingly heavy providing a nice feel. The VU meters are without any driver circuit, but do have about 4 diodes and a couple of resistors.
Regarding the VU meters, doing my homework I think Rod Elliot's project 55 would be fine, with the following modifications:
a) Prepend a buffer for isolation, as discussed earlier
b) Possibly build it dual mono to reduce possible crosstalk deterioration. Going that route, I would think of getting AC from each DCSTB for each channel, have a basic rectification with a full bridge and a voltage stabilizer. Then reference 0 volts of the VU meter board with the 0 volts of the respective DCSTB. Additionally feed volume remote control board and I-Select from the VU meter boards.
Does this make sense or is it a complete overkill? The transformers are 50VA per channel. Can i use them for the VU meter lights too or should I use a third TF?
Regarding the VU meters, doing my homework I think Rod Elliot's project 55 would be fine, with the following modifications:
a) Prepend a buffer for isolation, as discussed earlier
b) Possibly build it dual mono to reduce possible crosstalk deterioration. Going that route, I would think of getting AC from each DCSTB for each channel, have a basic rectification with a full bridge and a voltage stabilizer. Then reference 0 volts of the VU meter board with the 0 volts of the respective DCSTB. Additionally feed volume remote control board and I-Select from the VU meter boards.
Does this make sense or is it a complete overkill? The transformers are 50VA per channel. Can i use them for the VU meter lights too or should I use a third TF?
Ok, thanks. I didn't find the time to check the meters thoroughly, just a quick test. Measured 15MOhm resistance and lit using a 9V battery with 1k resistor in series. Could not light them with the DMM diode test or with an 1.5V battery. With 9V the light was intense enough, will try with 5-6V to see if I can get a lower luminosity. Hopefully will place the order today for the final missing parts to complete the build and start testing the actual circuit.
I visited a local headphones exhibition few days ago and I listened to some in a row. For closed back medium to elevated price category you were looking for I noticed a Japanese import model for sonics and nice wood sides. The Tago Studio T3-01. You might give it a look. No Dan Clark and Meze there but some nice Denons. The pricey Focal Stellia I found bit weird sounding.
For open back medium to elevated price category I noticed the Neumann NDH 30. The competitive Sennheiser HD 660S2 was clear but it did not impress me in musicality.