Tube DI Box

[Wavelength]

Generally.....PCB-fabrication shops can only hold their PCB board-outline dimensions to two (2) places (i.e., 3.93") UNLESS you want to pay extra for a "3-place held tolerance" on your board-outline and -- WHY -- would you even want to do that anyway???

/

That was a problem like 10 years ago. Now with 4mil traces and separation being the norm you can pretty much specify anything you want at no extra charge.

 

[Wavelength]

Ok since my AverTech is having issues with the new macOS, I ran some tests on the thing with the Jensen output on the Prism dScope III. I am going to order up a bunch of outputs including some UTM, Carnhill and others. Suggestions?
So here is the THD + noise, frequency response is above in a post. This looks pretty good and quiet considering it's strung out on my bench without a tube shield. Termination on the output was 600, Vrms output was ~1Vac, input as you can see was 0.22Vac rms gives it a gain of about 13dB.

 

[ruffrecords]

That's a lot of higher order harmonics at only 1vac output. That will definitely have a "sound".

Cheers

Ian

 

[MidnightArrakis]

That was a problem like 10 years ago. Now with 4mil traces and separation being the norm you can pretty much specify anything you want at no extra charge.

[you can pretty much specify anything you want at no extra charge] -- Maybe, but.....it actually boils down to which PCB-fabricator that you decide to use. I spent a couple of hours earlier this afternoon calling PCB-fabricators all over the U.S. here and their answers were all over the place when it came to how well they can hold their tolerances!!! Some stated that they typically route boards +/- 0.010" and +/- 0.005" will cost you extra, while others stated that they can typically route a board at 0.005" at no extra cost at all!!!

However, and this is similar for sheet-metal fabrication (although I am fairly certain your experiences are different), the PCB fabricators stated that they can hold a +/- 0.010" tolerance from the board-edge to the center-line of a mounting-hole and then a +/- 0.005" tolerance from that first mounting-hole to any other hole on the board. This is also how I dimension both my mechanical sheet-metal and PCB-design fabrication drawings.

/

 

[Wavelength]

That's a lot of higher order harmonics at only 1vac output. That will definitely have a "sound".

Cheers

Ian

Ian probably the highish output Z of the SRPP (~2300) is the reason for that into the 10K load. Still low thd but the drive isn't really as good. I have a folded cascode discrete DI board I was thinking of making up for a show I was doing in May in Florida (all acoustic) and never got a chance to finish it. I am going to make it up tomorrow so I can compare them.
Thanks,
Gordon

 

[Wavelength]

All finished up the board today by adding all the pads for the output transformer. I will order some via JLPCB soon.
I talked to David over at CineMag and he came up with a real nice 10K:600 PCB version. I ordered 11 of them in 50/50% M6/High Nickel.
I am going to do some front and back panels on a Hammond 1455 type chassis which is 4x4 and available in 6" and 8" long extrusions on Front Panel.
As above the new version has input Z at 1M, 500K and 100K and has an input TS and through TS with a male XLR. There is also a switch for termination at either 620 or 3K3 (of course you can change this to optimize to other outputs). There is also a switch for XLR PIN 1 to ground or open.
Thanks,
Gordon

 

[Wavelength]

Oh I almost forgot... I have been designing a number of tube DI for pedal boards as well. Maybe too many of them. I talked to a bass player friend this morning told him I am on probably design 164. Then talked to Jim Hagerman and had another idea. Just don't want the damn thing to draw a bunch of current. I have a 6418 clean boost with a HV supply and the whole thing only draws 35ma. Maybe something more robust than that one.