Demeter VTMP-2a with vibration-sensitive Tube PCB

[Helmholtz]

You might try a copper flux band on the PT to attenuate leakage. If well grounded it will also serve as an electrical shield.

[nevetslab]

The pair of 24-position 100k Step Attenuators arrived yesterday. This afternoon, I took to figuring out where the Input, Wiper and Ground terminals were, and upon finding them, I next set up to check the attenuation per step for the two units. The first step was 4dB, followed by around 3dB/step down to around -18dB. Then the steps changed a little before resuming around 3dB/Step, out to -50dB. Then increased to around 6dB/step out to shorting the wiper. I found an error on one, where the 'nominal' one was at -70dB and the other was at -44dB. Looked & measured the resistors, and found 33 ohms on one, with 680 ohms on the other at that step. So, repaired that one. Each step is made up of a 100k voltage divider, using Make-Before-Break switch steps, so each step stays at 100k total (like a pot), and each step has the set amount of attenuation. I've found another step that's 1/2 dB off, plus, I don't particularly like the first step being 4dB. So, I still have to make a couple more tweaks on these two attenuators before moving on to the installation. Apart from the other error I found, having a 1/2dB error, the nominal tracking error per step is less than 0.1dB/step.

The silk screen on the front panel is set for the traditational 270 or 300 deg arc between start and stop of the pots. These virtually go from start to stop, where the gap between those steps is the same as each attenuation step. Not yet sure if I can cobble something professional-looking to place onto the face plate to mark it up. If I still had Auto-Cad, I'd tackle it. I do have Solid Works 99, though I never did take the time to learn it. Perhaps this IS the time.

I also received the higher capacitance valued buss caps for the Heater supply, along with some 4A/400V Bridge Rectifiers for the same circuit. I'll get those installed tomorrow.

[nevetslab]

I revised the Attenuator's first step, so it's around -3dB instead of -4dB. Had to correct a step at -9db on one of them as well. Now, both are measuring the same thru the full range of attenuation. I set those aside after recording my notes.

I next removed the newly-replaced 1000uF/35V Buss cap following their 2A/50V bridge rectifier with a 4700uF/35V cap & 4A/400V bridge rectifier. I had to drill out the PCB mtg holes for the bridge, as they increased from 0.029" dia to 0.049" dia. Also had to apply a clamp to pull in the LM317K's folded heat sink fins adjacent to the buss cap to keep it from direct contact. I added a 10uF/35V filter cap on the output, along with a protection diode between the regulator output and the buss cap, just to be safe.

When I powered up the unit, headphones back on, fed from the Balanced Output to the Headphone Amp, I was no longer hearing ANY hum from the power transformer. When I dropped the AC Mains below 100VAC, I began hearing a slight filtered hum, though I didn't see the regulator drop out until 90VAC. At 120VAC, the Buss Voltage was now 21.15VDC; at 100VAC, it was down to 16.9VDC, and at 90VAC, where the regulator finally begin to drop, it was 15.0VDC.

Back to nominal line voltage, with phones on, no hint of xfmr hum at any gain setting (30dB, 40dB, 50dB), and just a fine sheen of wideband noise at any gain setting for both channels. That's a major improvement from what I had with the 1000uF buss cap! I checked it with my AKG C3000B mic I had on hand, and that too was a major improvement from what I was hearing before.

Now, to tackle installing these Step Attenuators. I'll post photos of the revisions, including the shock mount standoffs, elevated output xfmrs, etc.

[nevetslab]

I've got this completed. After cutting down the length of the attenuator shafts, being about a 1/4" too long for the Collet knobs, I then desoldered the cable leads of the two channels shielded cables wired to the 100k Audio Taper pots. As luck would have it, the leads were tinned & folded around the solder lugs prior to soldering,so had to take care in desoldering them not to cause the core insulators to melt. Got them removed, trimmed & tinned for the next step, then removed the two pots. I didn't bother with the nylon locating pin, but instead chopped those off, and used large 3/8" bushing ITL washers to lock them in place. 8mm threaded bushings on these attenuators. Installed them, then with the first collet knob in place, set it to 0dB and rotated the knob to align with first hash mark, just to see if the silk screen hash marks had any utility. I'll be horn-swaggled! The hash marks perfectly align with the 24 steps of the attenuators! So, I then rotated the knob to where 0dB would be, and using a small machinist scale, scribed three more hash marks between the ends of the silk screen to carry out the last 3 steps. Set the 0dB position, and checked again.

I next powered up the unit, to give a listen to the Attenuators to see if any shielding was going to be necessary around them. No problems. In the first few steps, you can hear slight soft clicks, then nothing as you continue turning the attenuation down. Both channels behave the same, and at 50dB gain, still no problems. Terriffic!

Then, having printed out the Attenuator Measurement results, I sat down with the Brother P-Touch and used Medium Size Text on 0.23" wide tape (Wht on Blk), and printed out the attenuation steps in 6dB steps. The steps deviate in a couple places out to -50dB, but, none of my doing. Now, having printed out the labels, cut them apart, and with an X-Acto knife, peeled the backing off and placed the labels onto the front panel around the two knobs.

Then, stopped to take another set of photos. I've attached the Attenuator Measurements Chart, along with the resistors used in building the step attenuators.

Step Attenuator Measurements.pdf
Demeter VTMP-2a Power Supply_Rev 3.pdf

[nevetslab]

The last couple days, I finally got around to measuring the difference in output drive capability of this preamp, now using 12AU7's in place of the taller 12BH7's. I already had found the plate current was a bit less....8mA for the 12AU7's vs 10.8mA for the 12BH7's. Driving open circuit (well, 100k input Z to my Amber 3501a Audio Analyzer) or 10k Bridging loads, there's little difference found. From open circuit max output of +27dBu (17.5V RMS) for the 12AU7's, and +27.5dBu (18.5V RMS) for the 12BH7's, and 0.3dB less driving 10k Bridging, that's insignificant. Driving 600 ohm loads (which from the units' xfmr output, causing a 4.1dB loss), the 12AU7 has a max output of +15.4dBu (4.5V RMS) vs +17.9dBu (6.1V RMS) for the 12BH7's, that's a credible difference, almost 3dB. Though, these days, how often are we driving 600 ohm loads? Granted, if doing so, it does reduce the headroom a little bit.

During this exercise, when I had the top cover open, I found the new 4A Bridge Rectifier was running VERY HOT, as was the 4700uF Buss Cap, sandwiched between the regulator heat sink and the bridge rectifier. When I checked the temperatures with my Fluke 80T-IR Infrared Temp Probe, the Bridge was 82 deg C, the Heat Sink was 74 deg C, and the Buss Cap was about 57deg C. Way too hot. I thought about that overnight, thinking there was a mounting hole thru the new 4A Bridge. There wasn't. So, after looking for a while for an available heat sink, not finding anything, plus lack of thru-hole, the thought of a workable solution with THAT bridge was fading fast. Why not mount a 25A Bridge to the side wall, and wire it up with quick-disconnects from the PCB? That made sense.

So, removed the side wall, punched a 0.188" dia hole thru the steel panel, reamed it out a bit more so a #10-32 screw could easily pass thru it. Greased up the 25A/400V Bridge Rectifier I had on hand, and mounted it to the panel. I had already lifted the PCB up to unsolder the bridge, and while I had it lifted up, recorded the output wire colors feeding the Preamp PCB from underneath the Power Supply PCB. Cut and prep'd discrete wires, keeping the same wire color to the bridge as used from the Power Xfmr. Set the PCB back down, installed the female 1/4" Fast-On Terminals, then mounted the side panel and plugged the bridge wires into place.

I was also able to open the twisted power supply harness to extract the RED-WHT heater supply wire, so I could clamp my current probe around it to measure the heater current. I powered the unit back up, clamped the current probe around the AC input to the bridge, as I had done earlier with the 4A bridge. Same 3.1A peak currents, measured the peak AC current INTO the Filter Cap/Regulator Circuit feeding the vacuum tube heaters, then moved the probe to measure the DC heater current flowing thru the output wire in the harness to the Preamp PCB.

3.1A Peak current into the 4700uF/35V Buss Cap (which has a max charging current rating of around 3700mA), and 0.80A DC current flowing into the heaters of the five tubes. Checked the temperatures after having it running for over half an hour. the regulator heat heat sink was 63.2 deg C, the Filter Cap was 42 deg C, and the side-wall mounted Bridge was 33.6 deg C. That's more like it!

Demeter VTMP-2a Power Supply_Rev B.pdf