Managed to build the borbelys for less than £50 partly because I had the valves. However ECC81s and ECC83s are readilly available. Problem comes with HT transformer, it needs 190VAC at 20mA and 10.8V at 1A and 10.8V at 2mA. Custom build is expensive. I used 5 x 1.6VA 0-22 + 0-22V Toroids for the low current supplies and a 30VA 0-15V for the heaters. All nice and neat as the transformers are tiny and all fit inside the enclosure. I'll post pictures once they arrive and it's working. At the moment I am using a custom built transformer using a 100VA transformer kit (which is F.UGLY)
Also, please post listening impression when using the
vis-a-vis the Borbelly if you have time. Thanks!
Also, please post listening impression when using the
Papa Borbely's bal tube line is decent one ....... to be more precise - elegant one , indeed .
but - to be able to compare it with
(equipped with Shunty) , one need equally well executed PSU stagesay that old geezer - Allen Wright have some interesting things in that area - Super Reg ......
Papa Borbely's bal tube line is decent one ....... to be more precise - elegant one , indeed .
but - to be able to compare it with
say that old geezer - Allen Wright have some interesting things in that area - Super Reg ......
Thanks for your comments, appreciate it. Let me start reading up on the
. I don't expect to learn much but should still be enough to ask stupid questions... 
try to expand building process longest you can
Glad I'm getting that bit right!
Well the Borbelly Valve Pre-Amp is almost finished and is being tested.
Although the initial results are good, I feel that the amplifier is adding a bit of timbre to the music. If that is the right phrase. Without the Borbelly the S's sound like S... With the Borbelly they sound like TSssss.
It's difficut to know if this is just because it is louder, as, when I remove the Borbelly and turn up the volume I also get TSsss.
Anyway, here's the photo that I promised.
It looks complicated but its not really.
The Borbelly uses slightly unusual power rails and didn't lend itself to a commonly available transformer. I didn't want to use an enormous case so that left me with a transformer problem.
At the left of the picture you will notice 5 x 1.6VA Mains Toroidals, each producing 22 + 22V AC. 5 of these secondaries in series gave me the 198V AC that I was looking for.
The Borbelly also needs a separate -12V DC supply for the Cathode Bias. The original intention was to use the 6th 22V secondary. However, in my quest for reliability, I decided to use a ladder PSU and the 1.6VA transformer did not have the current capacity for the relay. S**T, you will notice the extra E/I transformer at the front left.
The PCB just to the right of the tiny toroids is simply there to connect 5 transformers in parallel across the Mains and in series across the secondaries.
The PCB between the transformer PCB and the Borbellys is my Ladder PSU.
Valves don't like having 12.6V DC being applied to their heaters when cold. This is a common cause of valve failure.
Valves also suffer, all-be-it marginally, from cathode stripping if the HT is applied before the cathodes have warmed up.
So - the LADDER PSU.
1. At turn ON the -12V cathode bias is generated which starts a 20 second delay circuit to RL (Relay) 1, the HT relay.
2. At turn ON the +12.6V Heater Supply is gradually applied to the heaters, achieving full current in about 2 seconds.
3. 20 seconds later the 198V AC is applied to the CRCRC filter circuit for the HT supplies.
Hence, a close look at the PCB will show three distinct PSUs. The 6 x 150uF capacitors at the top at for the HT, three for each channel. The little bit of VEROBOARD is the slow start for the HT. The larger heatsink accomodates the LM317 which is regulating the heater supplies.
I forgot to mention the heater transformer. This is the third transformer, the larger Toroid amongst the others.
As I'm using a 15V 30VA Toroid, the LM317 was getting a little bit HOT, so to help it out I placed a 5W 1R0 resistor in series with the transformer secondary.
Careful choice of the components now allows the whole unit to fit in a 1U case with only the valves poking through. I haven't drilled the case yet as I'm still convinced that I like this amplifier. I might wait until the Pumpkins are ready before I sacrifice a £60 enclosure.
Although the initial results are good, I feel that the amplifier is adding a bit of timbre to the music. If that is the right phrase. Without the Borbelly the S's sound like S... With the Borbelly they sound like TSssss.
It's difficut to know if this is just because it is louder, as, when I remove the Borbelly and turn up the volume I also get TSsss.
Anyway, here's the photo that I promised.
An externally hosted image should be here but it was not working when we last tested it.
It looks complicated but its not really.
The Borbelly uses slightly unusual power rails and didn't lend itself to a commonly available transformer. I didn't want to use an enormous case so that left me with a transformer problem.
At the left of the picture you will notice 5 x 1.6VA Mains Toroidals, each producing 22 + 22V AC. 5 of these secondaries in series gave me the 198V AC that I was looking for.
The Borbelly also needs a separate -12V DC supply for the Cathode Bias. The original intention was to use the 6th 22V secondary. However, in my quest for reliability, I decided to use a ladder PSU and the 1.6VA transformer did not have the current capacity for the relay. S**T, you will notice the extra E/I transformer at the front left.
The PCB just to the right of the tiny toroids is simply there to connect 5 transformers in parallel across the Mains and in series across the secondaries.
The PCB between the transformer PCB and the Borbellys is my Ladder PSU.
Valves don't like having 12.6V DC being applied to their heaters when cold. This is a common cause of valve failure.
Valves also suffer, all-be-it marginally, from cathode stripping if the HT is applied before the cathodes have warmed up.
So - the LADDER PSU.
1. At turn ON the -12V cathode bias is generated which starts a 20 second delay circuit to RL (Relay) 1, the HT relay.
2. At turn ON the +12.6V Heater Supply is gradually applied to the heaters, achieving full current in about 2 seconds.
3. 20 seconds later the 198V AC is applied to the CRCRC filter circuit for the HT supplies.
Hence, a close look at the PCB will show three distinct PSUs. The 6 x 150uF capacitors at the top at for the HT, three for each channel. The little bit of VEROBOARD is the slow start for the HT. The larger heatsink accomodates the LM317 which is regulating the heater supplies.
I forgot to mention the heater transformer. This is the third transformer, the larger Toroid amongst the others.
As I'm using a 15V 30VA Toroid, the LM317 was getting a little bit HOT, so to help it out I placed a 5W 1R0 resistor in series with the transformer secondary.
Careful choice of the components now allows the whole unit to fit in a 1U case with only the valves poking through. I haven't drilled the case yet as I'm still convinced that I like this amplifier. I might wait until the Pumpkins are ready before I sacrifice a £60 enclosure.
both cases - not so much of an issue , as much one can have impression reading fairy tales on net
I agree. A lot of valve amplifiers apply all power supplies simultaneously on power-on.
6.3V Heaters are not too bad, but even so the cold inrush current can cause stress to the heater windings. 12.6V heaters (as in the ECC81 and ECC83 in the Borbelly) draw significant cold current which can result in heater failure.
A simple SLOW-START using an LM317 is neither difficult nor expensive and just gives yo the peace of mind that you are treating the valves gently.
Even with the SLOW-START circuit the heaters glow almost yellow on initial application of power. I've monitored the voltage applied and it starts at about 1.2V climbing to 12.6V in around 2 seconds.
Cold Cathode stripping is a bit of a myth as you suggest, but what the hell. One transistor and a relay wont break the bank.
I'm just collecting the components together for the Pumpkin as we speak.
6.3V Heaters are not too bad, but even so the cold inrush current can cause stress to the heater windings. 12.6V heaters (as in the ECC81 and ECC83 in the Borbelly) draw significant cold current which can result in heater failure.
A simple SLOW-START using an LM317 is neither difficult nor expensive and just gives yo the peace of mind that you are treating the valves gently.
Even with the SLOW-START circuit the heaters glow almost yellow on initial application of power. I've monitored the voltage applied and it starts at about 1.2V climbing to 12.6V in around 2 seconds.
Cold Cathode stripping is a bit of a myth as you suggest, but what the hell. One transistor and a relay wont break the bank.
I'm just collecting the components together for the Pumpkin as we speak.
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Finally the Borbely pre-amp is in its almost completed state, all that is remaining is the front panel to make a cut-out for the ON/OFF switch and make a hole for the LED indicator.
For the spotters amongst us - yes I include myself - you will notice that component count on PCBs does not match the PSU PCBs. This is simply because the Mod. 1 PCBs were slightly different and I haven't removed redundant components.
These pre-amps are lovely to listen to, inherently stable and easy to build. No awkward setting up to do. I wouldn't recommend them to an absolute beginner but certainly to someone with only limited experience.
Note you will need access to an Oscilloscope or an Audio Millivoltmeter to set them up. They can be operated without being set-up and adjust the settings according to your ears if you so desire.
I can't publish the Borbely Pre-Amp PCB schematics as they are copywright. However, they are available on-line and it is a simple job (using Microsoft Paint) to produce the artwork that you require. Any probs, just e-mail me.
Incidentally, I included 2 x changeover switches at the output XLR connectors.
I have yet to prove what I have read in these forums, but I have heard that occasionally CDs are recorded with the channels in the incorrect phase.
As it was simple to implement, I included the ability to change the phase of one or both channels in order to correct this.
I have yet to prove what I have read in these forums, but I have heard that occasionally CDs are recorded with the channels in the incorrect phase.
As it was simple to implement, I included the ability to change the phase of one or both channels in order to correct this.
I like to use LM317 as CCS for heaters ; just set exact voltage across heated filament (meaning - use current from datasheet as starting point and finally set voltage ) - for E line of tubes
that way you're also having best soft start you can imagine - (say) 300mA from cold to hot filament .
for P line of tubes , choice is even easier ; take PCC88 for instance - just set current to 300mA .......
that way you're also having best soft start you can imagine - (say) 300mA from cold to hot filament .
for P line of tubes , choice is even easier ; take PCC88 for instance - just set current to 300mA .......
I can see your point. I've got 6 x heaters at approximately 150mA each. Let the LM317 do the work and keep the curent constant.
It would be interesting to see the LM317 in CCS Slow Start mode so that even then the current is limited at switch ON. I'm open for suggestions.