Has anyone else been brave enough to build one of these leviathons from our Chinese friends at DIY-GENE ?
I've built a Dual Mono version using the case that DIY-Gene advertise on E-Bay. Unfortunately my Mk.1 only worked for a couple of months before eating one channel.
This in part may be my fault. I have opted for 500VA transformers at 45-0-45V. This results in the PSU rails sitting at +/- 65VDC instead of the +/- 62V that is specified. During construction I used the supplied Sil-Pad insulating washers for the output transistors, coupled with the extra supply voltage, I guess that the devices got a bit hot and eventually "popped".
Mk.2 has been assembled with mica washers and very meagre quantities of thermal grease between ALL thermal junctions.
DIY-GENE were not forthcoming with the setting-up procedure for the repaired boards so I've had to use a bit of engineering guesswork.
The original amplifiers were biassed at 1.25A, however, this results in the repaired amplifier heatsinks running at just under 70 degrees which seems excessive. I have throttled them down to 1.0A and now they run at about 55 - 60 degrees (ambient = 22 degrees).
DIY-GENE sell this kit as a MarHevison A700. It is supposed to be a clone of a Mark Levinson - HAS ANYONE GOT ANY IDEA WHICH DESIGN IT IS A CLONE OF ?
I'm thinking about attempting an Aleph 4. Is the Aleph significantly better than the Mark Levinson Clone ?
I'm using B&W DM602 s3's with an Arcam CD36 front end. Chord Chorus Interconnects and QED Silver Speaker cables.
I've built a Dual Mono version using the case that DIY-Gene advertise on E-Bay. Unfortunately my Mk.1 only worked for a couple of months before eating one channel.
This in part may be my fault. I have opted for 500VA transformers at 45-0-45V. This results in the PSU rails sitting at +/- 65VDC instead of the +/- 62V that is specified. During construction I used the supplied Sil-Pad insulating washers for the output transistors, coupled with the extra supply voltage, I guess that the devices got a bit hot and eventually "popped".
Mk.2 has been assembled with mica washers and very meagre quantities of thermal grease between ALL thermal junctions.
DIY-GENE were not forthcoming with the setting-up procedure for the repaired boards so I've had to use a bit of engineering guesswork.
The original amplifiers were biassed at 1.25A, however, this results in the repaired amplifier heatsinks running at just under 70 degrees which seems excessive. I have throttled them down to 1.0A and now they run at about 55 - 60 degrees (ambient = 22 degrees).
DIY-GENE sell this kit as a MarHevison A700. It is supposed to be a clone of a Mark Levinson - HAS ANYONE GOT ANY IDEA WHICH DESIGN IT IS A CLONE OF ?
I'm thinking about attempting an Aleph 4. Is the Aleph significantly better than the Mark Levinson Clone ?
I'm using B&W DM602 s3's with an Arcam CD36 front end. Chord Chorus Interconnects and QED Silver Speaker cables.
An externally hosted image should be here but it was not working when we last tested it.
This is the schematic:Hi,
a 45-0-45Vac transformer and ClassA cannot be, surely!
1.25A of bias and that transformer cannot be ClassA bias into any normal speaker load.
2.5Apk of ClassA into an 8r0 load would be ~25W
But an amplifier that is running on +-65Vdc supply rails might be expected to give ~210W of ClassAB output into 8r0.
This seems much more like ClassAB bias.
ps.
your link doesn't.
a 45-0-45Vac transformer and ClassA cannot be, surely!
1.25A of bias and that transformer cannot be ClassA bias into any normal speaker load.
2.5Apk of ClassA into an 8r0 load would be ~25W
But an amplifier that is running on +-65Vdc supply rails might be expected to give ~210W of ClassAB output into 8r0.
This seems much more like ClassAB bias.
ps.
your link doesn't.
Last edited:
I tend to agree with your maths. The Alephs are biassed at about 3A plus. This is why I ask the question, what is this supposed to be a clone of. The sound is rather superb though, when compared with my Arcam AVR-280, allbeit it a heavilly modified AVR.
Sorry that the schematic doesn't seem to come out. If you search on DIY-GENE then go to schematics, you can download the A700 Circuit Diagrams.
Sorry that the schematic doesn't seem to come out. If you search on DIY-GENE then go to schematics, you can download the A700 Circuit Diagrams.
Under the 'Scope
My maths gives me a slightly higher figure.
P = I squared x R, the DM602s are about 6 ohms - most of the time.
If Iq is set at 1.25A then Ipk = 2.5A. I squared then becomes 6.25
6.25 x 6 is just over 36 Watts.
Using an oscilloscope the amp is delivering about 60 Watts into a 10 Ohm dummy load. The current drawn from the power supply stays constant at 1.25A, the meters flicker occasionally on particularly loud transients but everything tends to indicate that the amplifier is running in Class A most of the time.
The Alephs that I have seen have all had substantially more heatsinking than this beast. This is why I am interested in finding anyone else who has built one.
My maths gives me a slightly higher figure.
P = I squared x R, the DM602s are about 6 ohms - most of the time.
If Iq is set at 1.25A then Ipk = 2.5A. I squared then becomes 6.25
6.25 x 6 is just over 36 Watts.
Using an oscilloscope the amp is delivering about 60 Watts into a 10 Ohm dummy load. The current drawn from the power supply stays constant at 1.25A, the meters flicker occasionally on particularly loud transients but everything tends to indicate that the amplifier is running in Class A most of the time.
The Alephs that I have seen have all had substantially more heatsinking than this beast. This is why I am interested in finding anyone else who has built one.
absolutely not.
At quiescent condition each supply rail will deliver Ibias of the output plus I quiescent of the voltage amplifier stage.
The output stage draws varying current from both supply rails as soon as it starts to deliver current to the output.
You cannot have a constant 1.25A from the supply rails if it delivers current to the load
The lack of variation of supply rail current cannot differentiate ClassA from ClassAB. Both have a constant quiescent current. Both have variable currents on both supply rails when the load receives current.
When the 60W is being delivered to the 10r load the output is 34.64Vpk and 3.464Apk on the resistive load.
What are the average and minimum voltage on each supply rail while delivering this output current
Last edited:
With 500VA transformers (1 each channel) and 0.3F (300mF / 300000uF) of reservoirs on each of the 4 supply rails the supply rails are absolutely ROCK SOLID at even the most ear bending volumes. At FULL POWER the supply rails are +/- 65V with aboout 20mV of ripple (probably caused by the Bridge Rectifiers).
Strangely, this ripple is audible, allbeit VERY QUIET. I have to put my ears almost on the speaker cones to hear it.
Comparing my construction with Aleph projects, my heatsinks are quite small. They are 160mm high x 420mm long and 50mm deep with finned vanes.
Strangely, this ripple is audible, allbeit VERY QUIET. I have to put my ears almost on the speaker cones to hear it.
Comparing my construction with Aleph projects, my heatsinks are quite small. They are 160mm high x 420mm long and 50mm deep with finned vanes.
Is it possible to attach images here without a URL. I don't tend to leave my computers on all the time so a URL link wont always work.
In my job we have a saying, "A fool is someone who says he knows everything"
TOO TRUE, after 40 years in this business I am still learning.
I appreciated that DVMs may fail to indicate instantaneous currents due to their averaging nature. But even with an analogue meter in the supply chain I wasn't seeing massive current demands at FULL POWER. Everything just seems to sit quite happily at 1.25A.
As my anolgue mulimeter is quite old (ancient actually), I tried putting 1R0 resistors in the supply rails and measuring the current with the 'scope. The results were the same.
TOO TRUE, after 40 years in this business I am still learning.
I appreciated that DVMs may fail to indicate instantaneous currents due to their averaging nature. But even with an analogue meter in the supply chain I wasn't seeing massive current demands at FULL POWER. Everything just seems to sit quite happily at 1.25A.
As my anolgue mulimeter is quite old (ancient actually), I tried putting 1R0 resistors in the supply rails and measuring the current with the 'scope. The results were the same.
Why would you buy from people having no respect for trademarks an IP whatsoever, and, in the meantime promoting them when seeking the help they can't provide?
It is those kind of people that kill diy.
/Hugo
- Status
- Not open for further replies.