I believe it's the Magnequest EXO-99 with a slightly different gap
You can get almost 600H in series and 160 in parallel
Custom is the best way to match it up with the tubes youre using.
The output impedance depends on where your AVC volume is set. It could easily range from 500-4K depending on listening levels.
Last edited:
I guess, you didn't understand correctly:
about 5k is the choke loaded 01a tube's output impedance, aka generator impedance before capacitor coupled inductor (AVC).
Each TVC/AVC measurement suggest 50R as generator impedance, and in practical not more than few hundred Ohm.
If you use CCS, or Ale's gyrator as load of 01a, the output impedance range is -about- hundred...few hundred Ohm. In case of FET source follower it is few ten Ohm.
about 5k is the choke loaded 01a tube's output impedance, aka generator impedance before capacitor coupled inductor (AVC).
Each TVC/AVC measurement suggest 50R as generator impedance, and in practical not more than few hundred Ohm.
If you use CCS, or Ale's gyrator as load of 01a, the output impedance range is -about- hundred...few hundred Ohm. In case of FET source follower it is few ten Ohm.
We had a 9 preamp shootout today with Ale Moglia and Tony Rees, with one very significant result which I, for one, didn't expect. This was an 01A preamp (gyrator + source follower) that Ale had put together the day before on a hunch with 6 Cree C3D02060F diodes in the cathode. It was not only better than the other 10Y, 01A, EML20A, 6E6P-DR and 2P29L preamps (most in filament bias), it was comprehensively better in all ways. More transparent, better instrumental timbre, cleaner and all round more engaging and a better listen. If this can be reproduced for a variety of DHTs it marks a compete change in my builds from filament bias to SIC bias. I can't wait to get the soldering iron out - I want this sound in my system!
I'm sure Ale will be writing this up in more detail in his blog and with photos, so this is just advanced notice of what's coming!
I'm sure Ale will be writing this up in more detail in his blog and with photos, so this is just advanced notice of what's coming!
Cree are also very good as B+ rectifier. Do you know how many will be necessary for 26 preamp?
Hi there,
I may add that I used to use these SIC Diodes first on a 4P1L PreAmp as filament resistor replacement with excellent results. That motivated to use them also on a 01A PreAmp with very convincing acoustic results. I finally dropped Ale's (very fine) gyrator in favour of a 600H (I guess 400H should do) choke which gave me a more refined top end. I furthermore found operating the filament power supply in C-L-C mode with a low DCR choke relevant.
My current 01A Pre also features a SIC source follower driving a RS1003 power amp.
All the best,
Ulrich
I may add that I used to use these SIC Diodes first on a 4P1L PreAmp as filament resistor replacement with excellent results. That motivated to use them also on a 01A PreAmp with very convincing acoustic results. I finally dropped Ale's (very fine) gyrator in favour of a 600H (I guess 400H should do) choke which gave me a more refined top end. I furthermore found operating the filament power supply in C-L-C mode with a low DCR choke relevant.
My current 01A Pre also features a SIC source follower driving a RS1003 power amp.
All the best,
Ulrich
Sorry for speaking that low.
On the 4P1L pre I used 8 Cree C3D02060 in series as the filament "resistor" to get the gate voltage of (I need to work on my documentation) -8.4 (?) Volt.
Regarding your second question, i.e. the max. number of SiC diodes in series before the acoustic result might get less good: I can't answer.
I THINK that their successful application is partially due to the low dynamic resistance of these diodes AND -as they do not change their resistance with current variations a lot (and less than any resistor)- this seems to comes close to a fixed gate voltage schematic.
I can furthermore second "merlin el margo"'s statement of SiC diodes in the B+ power supply, compared to other diodes. I did not do any comparison to tube rectifiers, except -due to their optics- 2 mercury vapour rectifiers (AX50, 394A). Initially they sounded "more interesting", but finally I found that this was caused by their HF noise spreading everywhere. Therefore I stopped using them.
All the best,
Ulrich
On the 4P1L pre I used 8 Cree C3D02060 in series as the filament "resistor" to get the gate voltage of (I need to work on my documentation) -8.4 (?) Volt.
Regarding your second question, i.e. the max. number of SiC diodes in series before the acoustic result might get less good: I can't answer.
I THINK that their successful application is partially due to the low dynamic resistance of these diodes AND -as they do not change their resistance with current variations a lot (and less than any resistor)- this seems to comes close to a fixed gate voltage schematic.
I can furthermore second "merlin el margo"'s statement of SiC diodes in the B+ power supply, compared to other diodes. I did not do any comparison to tube rectifiers, except -due to their optics- 2 mercury vapour rectifiers (AX50, 394A). Initially they sounded "more interesting", but finally I found that this was caused by their HF noise spreading everywhere. Therefore I stopped using them.
All the best,
Ulrich
Dear Felipe,
yes I do operate a heater set-up similar to the Coleman boards, i.e. the tube filament is clamped between a gyrator and a CCS, based on the initial Coleman postings.
To replace the filament resistor you should look up the CREE C3D02060 data sheet to see what the forward voltage drop will be for what current for ONE diode. Based on this value you'll know how many diodes you'll need.
In my case it had been 1.05V per diode (@ about 600mA), so 8 diodes for a gate voltage of -8.4V.
It may be a good idea to verify the estimated gate voltage in practice by simply applying the requested current through a string of these diodes outside the amp before implementing.
All the best,
Ulrich
yes I do operate a heater set-up similar to the Coleman boards, i.e. the tube filament is clamped between a gyrator and a CCS, based on the initial Coleman postings.
To replace the filament resistor you should look up the CREE C3D02060 data sheet to see what the forward voltage drop will be for what current for ONE diode. Based on this value you'll know how many diodes you'll need.
In my case it had been 1.05V per diode (@ about 600mA), so 8 diodes for a gate voltage of -8.4V.
It may be a good idea to verify the estimated gate voltage in practice by simply applying the requested current through a string of these diodes outside the amp before implementing.
All the best,
Ulrich
