A case and power supply for Victor's low THD oscillators

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks guys for the heads up on the backwards battery snaps. I did indeed notice it before anything exciting happened. I wanted to post my findings though, I have a few head scratchers to also ask about, which are potentially related to my analyzer output.

To test, I let things rest for an hour on AC power with the three oscillators turned on. The negative input to the analyzer is terminated with a 75R BNC jobbie. I'm posting results from both channels, AC and battery power, across three frequencies.

First, I tend to get some very wacky THD readings. See 10kHz_R, 11kHz_R_Batt and 11kHz_R. Since the graphs look great, I assume this is just an analyzer quirk? In other graphs, it shows pretty close to 0 and sometimes "---" which I presume means, nearly zero?

I see some 60Hz coming through from the analyzer power. However, it's not consistently there!

In general, do things look about as you would expect them?

edit: looks like the forum looses the original file names. in order, left-to-right, they are:

1kHz_L_Bat
1kHz_L
1kHz_R_Bat
1kHz_R
10kHz_L_Bat
10kHz_L
10kHz_R_Bat
10kHz_R
 

Attachments

  • 1kHz_L_Bat.png
    1kHz_L_Bat.png
    40.2 KB · Views: 550
  • 10kHz_R.png
    10kHz_R.png
    42.8 KB · Views: 106
  • 10kHz_R_Batt.png
    10kHz_R_Batt.png
    42.7 KB · Views: 94
  • 10kHz_L.png
    10kHz_L.png
    38.3 KB · Views: 112
  • 10kHz_L_Batt.png
    10kHz_L_Batt.png
    38.3 KB · Views: 520
  • 1kHz_R.png
    1kHz_R.png
    43.7 KB · Views: 519
  • 1kHz_R_Bat.png
    1kHz_R_Bat.png
    38.5 KB · Views: 533
  • 1kHz_L.png
    1kHz_L.png
    43.3 KB · Views: 546
Last edited:
Member
Joined 2004
Paid Member
The 10 KHz and 11 KHz harmonics will all be outside of your measuring range. You need to switch to 192KHz to see any harmonics from them. If you mix the two you can then see the IM products which should be 1 KHz, 9KHz & 12 KHz to see in band with the 48 KHz sample rate. The peak value will be significantly higher than the RMS value so the reference levels may be misleading.

I think the first picture is on AC since it has a 60 Hz spur. The lack of spurs on battery is pretty remarkable and attests to Quantasylum's layout. The harmonics at around -120 dBC is what I would expect from that ADC.

I don't know what the actual levels are but if you calculate -120 dB from FS 2V thats 2 uV. The source impedance of the oscillator is 600 Ohms. So to get 2 uV across 600 Ohms you need (I think) 30 nA (too many zeros. . . .) Then calculate the capacitor needed to get 30 nA from 120V is something like 1 pF at 60Hz. Again too many zeros so I may be an order of magnitude or more off but the point is that it takes very little leakage to get those low level spuria. I usually spend more time moving ground connections and wires around the bench than taking measurements.
 
Back, updated the project materials

Sorry for disappearing for (long) while! Day stuff has been really busy and interfering with the audio hobby. :) In fact, looks like pretty much one year ago almost exactly when I last was able to spend time with this project.

Several people have PMed me about Victor boards over the past few months and I haven't been able to reply. I think I've gotten back to everyone now over the last week. If I've missed anyone, please send a PM.

The board is now a "kit" of sorts. The comparator chip is pre-soldered since it is surface mount and tiny, the only package it comes in. The rest of the board is relatively easy soldering, most of it through hole. Also included are the screws to bolt the board to the Hammond case, bolts and nuts for attaching the CCS heat sinks, nylon standoffs for the board corners while soldering, and a bolt and nut for attaching the terminal strip to the front panel. $28 for the board kit, plus shipping which is $3.50 in the US and $14 international (actual cost, our post office sucks for international), plus paypal. Paypal is 2.9% + $0.50 for us, 3.9% + $0.60 for international.

I don't sell parts kits, but all the Mouser numbers and many Digikey numbers are in the BOM. I should have all the parts here, so if Mouser or Digikey is out of something, including the alternates I've added to the BOM, let me know. I probably have it and would sell out-of-stock parts like that.

I've updated the BOM and posted on the project Google Drive link (the latest board is still V3.2)

V 3.1 & V 3.2 - Google Drive

Several things needed updating. TDK has changed the part numbering on their MLCC caps over the last year. The On Semi 1N4004 rectifiers are on order everywhere, so I've specified alternates. The main filter caps have a new and improved version. What Mouser has in stock for resistors bounces around, so current stock shown. I've updated the Mouser and Digikey prices on everything. For the few things that are "on order" at both I've listed alternate part numbers.

And some exciting new stuff - Victor has a new version of his board out! The pot is now a Bourns and is in-line with the RCA jacks. The ground screw is different now. Before, in my original design, Victor has a nut soldered to the bottom of the board. Now The bolt is in-line with the RCA, between them, and a washer and nut are used to press the back of those jacks forward. A big thanks to Victor in sending me the latest mechanical diagram and photos.

I have the front panel CAD for Front Panel Express (and PDF images of each) revised now and posted at the link above to match Victors' new boards. BUT...a word of caution here... there are now a few versions of Victor's oscillator board. The X and Y (and Z, height) dimensions of Victor's oscillator board itself are the same with all, but where the front panel holes go changes. When it comes to the front panel, it is the user's responsibility to determine which of Victor's versions is being used in the (up to 4) Victor oscillator board positions, and modify the CAD file accordingly. The previous CAD files for Victor's original oscillator board are still posted. You may wind up with a mix of different versions of Victor's boards if you are mounting more than one, each with a different front panel hole pattern. There was another Version last year of Victor's oscillator that had the ground nut on the bottom moved up slightly from the original. Best thing to do is consult the Victor mechanical diagrams posted for the different versions and use a caliper (cheap at Harbor Freight tools here, like $5) to verify your board matches the mech diagram and associated CAD file you are using for the front panel.

And please note - the different versions of Victor's oscillator board does not have any effect on the Power Supply board here. Works with all. It is just the specifics of where the holes go in the front panel CAD that change when mounting his boards to the front panel. The rear panel CAD of the Hammond case is also the same for all.

Happy soldering! :)
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.