Tube DI Box
- Thread starter ruffrecords
- Start date
Help Support GroupDIY Audio Forum:
Wavelength
Well-known member
Ian,
By better I meant less losses.
Really with that much loss the design is not going to sound much different than a SS or passive version.
By better I meant less losses.
Really with that much loss the design is not going to sound much different than a SS or passive version.
C.B. - Boudio
Well-known member
Also interesting: EABC80's triode plate voltage in '50s radio's could be as low as around 50V and current 0,5mA. So, maybe it's can be possible to make a cathode follower work from phantom power with valves like those.
ruffrecords
Well-known member
ruffrecords
Well-known member
I have simulated a White follower version which seems to work rather well:
R2 is optimised (R2) as per Broskie but using the actual operating point rather than raw data-sheet values. Further optimisation is possible taking into account the relatively fixed load but distortion at +20dBu input is only 0.1% so it is probably not worth it. The output assumes one of my standard 2400:600 transformers so there is plenty of choice here (Carnhill, Edcor, OEP etc). The 1K load (R8) represents a 1K log output level control and R9 represent a typical mic input impedance. In parallel these represent the worst case load of 600 ohms. With a 0dBU input the distortion is a tad over 0.01%.
Cheers
Ian
R2 is optimised (R2) as per Broskie but using the actual operating point rather than raw data-sheet values. Further optimisation is possible taking into account the relatively fixed load but distortion at +20dBu input is only 0.1% so it is probably not worth it. The output assumes one of my standard 2400:600 transformers so there is plenty of choice here (Carnhill, Edcor, OEP etc). The 1K load (R8) represents a 1K log output level control and R9 represent a typical mic input impedance. In parallel these represent the worst case load of 600 ohms. With a 0dBU input the distortion is a tad over 0.01%.
Cheers
Ian
Last edited:
Wavelength
Well-known member
Ian,
~~~
White-Follower: If we are still using the Nixie tube supply the cathode in the top to filament voltage is in violation. Probably not a big deal. I take it you are looking at these pages:
https://www.tubecad.com/2019/04/blog0462.htm
I have never been really happy with the WF. I have used the SRPP in a number of amps including my Cardinal which won a number of awards. Even now I would rather listen to a resistor or inductor loaded triode over either the WF or an SRPP.
I had some ideas today, maybe when I get to work in the morning I could throw some out.
The signal will take on tube attributes of sound with the amount of gain. This is why EF86 and 12AX7A alter the sonic character so much.
~~~
White-Follower: If we are still using the Nixie tube supply the cathode in the top to filament voltage is in violation. Probably not a big deal. I take it you are looking at these pages:
https://www.tubecad.com/2019/04/blog0462.htm
I have never been really happy with the WF. I have used the SRPP in a number of amps including my Cardinal which won a number of awards. Even now I would rather listen to a resistor or inductor loaded triode over either the WF or an SRPP.
I had some ideas today, maybe when I get to work in the morning I could throw some out.
ruffrecords
Well-known member
Violation of what? If you are talking heater/cathode voltage then with a 180V supply we only have 90V on the cathode with Vhk rated at 130V dc. Or did you mean something else?
No, not seen that one. There are a number of papers that refer to Broskie in terms of optimisation. That is all I looked at.
Cheers
Ian
ruffrecords
Well-known member
I got a pair of these in the post this morning. Unfortunately there is no data sheet with them. Is the power in socket centre is positive?
I notice it is two separate SMPS using two different chips. - a MAX1771 for the HT and a CX8509 for the heaters. Both chips run in the 300KHz region. Let's hope there are no beat frequencies in the audio band.
Edit: Oh dear; I just discovered regular stomp box dc supplies are 9V with centre negative. Now I am really confused. I might just have to trace the PCB.
Cheers
Ian
Last edited:
rock soderstrom
Tour de France
Distortion, bandwidth and output impedance - are these the main (only) concerns designing the DI? Is there much difference in the output impedance between synths, guitars, basses and other things that we might DI that need to be considered? I feel like I'm missing something? Drive capability?
ruffrecords
Well-known member
There is a whole bunch of things to consider, some of which are pretty common and easy to specify and others that are less well defined.
Probably one of the biggest variables is the output impedance of the various devices that might be connected to its input. Guitars typically have pretty high output impedances and it usually varies a lot with frequency due to the electromagnetic properties of the pickups. Synths, on the other hand have relatively low output impedances that remain more or less constant. The common solution to this is to give the DI box an input impedance in the region of 1 Megohm (which is much the same as that of a tube guitar amp input). This happens to perfectly OK for other sources like synths.
Drive capability of the DI itself is being considered in detail as can be seen from the sims I have undertaken.
Perhaps the most important parameter is the one that is most difficult to quantify - what does it sound like? If you have seen any of my other designs you will know that I do not aim for 'tube sound' but wide bandwidth, low noise, low distortion and high headroom. In achieving that, the level of distortion is still much higher than the 0.0000000x% the op amps claim to give you, but for some reason a lot of people like the way they sound so that is what I shall continue to do.
Cheers
Ian
ruffrecords
Well-known member
Blast, I just ordered a regular stomp box style negative centre type. I will have to get a positive centre one as well.
Cheers
Ian
You could put a modified quick connect so you can swap polarity too.
TwentyTrees
Well-known member
Our bassist had one of those on his DI for live shows. Then one day it fell off, he didn't notice, and the DI went up in smoke at soundcheck! Good times.You could put a modified quick connect so you can swap polarity too.
merlin
Well-known member
- Joined
- Aug 17, 2011
- Messages
- 368
DI is really only needed for passive pickups, which are all high impedance. Around 1Meg input impedance is standard, although piezos may need more. Most other instruments are active and can be plugged straight into a line input with no need for DI.
If WCF is being considered then this seems like a good time to revisit the cathodyne unity-gain line driver
https://groupdiy.com/threads/cathodyne-mu-follower-line-out.86934/post-1139274
Last edited:
Wavelength
Well-known member
Ian,
Just cut the cord and solder to the two terminals on the right. Or do what I do and buy a bunch of the converters from Digikey. Think they are in the adapter between series.
~~~
Merlin, I was thinking choke loaded ECC88 (doing that now for a audio preamp) and then parallel feed the output with a 10K:600. Probably good for guitar and bass with a 0.47uF film cap. I am using a low nickel 150H/5ma on my audio design. Output is a 15K:500 peerless design. I have some wound in %99.999 silver with teflon insulation.
Just cut the cord and solder to the two terminals on the right. Or do what I do and buy a bunch of the converters from Digikey. Think they are in the adapter between series.
~~~
Merlin, I was thinking choke loaded ECC88 (doing that now for a audio preamp) and then parallel feed the output with a 10K:600. Probably good for guitar and bass with a 0.47uF film cap. I am using a low nickel 150H/5ma on my audio design. Output is a 15K:500 peerless design. I have some wound in %99.999 silver with teflon insulation.
ruffrecords
Well-known member
I got a couple of these in the post the other day and finally got round to trying one out this evening. The HT supply works well and powers up easily to 250V into a 20K load (12.5mA). However, I am having problems with the 6.3V heater output. It will only power up a single ECC88 heater (300mA). It will not power up two in parallel. I don't know of this is something to do with th wall wart I am powering it with but is is a 12V type rated at 1 amp which I would have thought was more than enough. Any ideas?
Cheers
Ian
rock soderstrom
Tour de France
I had no problem running 600mA for two tubes in parallel. But my wall wart has 12V/ 2A current capability, maybe that makes the difference?
Im not sure I understand the need for a board that produces ht and lt supplies ,
your obviously going to have a low volt DC supply rail , which can have a linear regulator to supply the heaters ,why complicate things with two chips that opperate at a different frequency ?
your obviously going to have a low volt DC supply rail , which can have a linear regulator to supply the heaters ,why complicate things with two chips that opperate at a different frequency ?
ruffrecords
Well-known member
Not so much a complication, just an alternative. Both chips work around 300KHz. Nothing is yet cast in concrete.
Cheers
ian
Similar threads
