uTracer (Micro Curve Tracer)

[martin manning]

Thanks, Mark. Started yours yet? It must be winter up there by now ;^)

[M Fowler]

No I haven't started on mine yet, too many guitar amp repairs, mods, radio repair, siding son's house, removing trailer house off a lot, cutting trees, and my own crazy stuff.

Mark

[martin manning]

Here is a zoomed-in view of plate current and grid current for positive grid voltage (to +1.5V) for a 12AX7 (The right-hand vertical axis is actually Ig, as measured by the screen channel). The base level of ~180µA seen at all negative Vg's is still an issue... If I trace a resistor connected between g1 and k, that offset is present at Vg-k=0, so I'm taking that to be an error that needs to be eliminated somehow. No ideas yet, though. Setting that aside, there is no increase in Ig until Vg-k reaches zero, at which point it rises about 20µA, and curls up a bit as Va-k gets very low, say < 40V. The Ig remains essentially constant at higher Va-k, up to 400V. I recall reading that grid current begins as Vg-k approaches zero, between -0.5 Vg-k and zero. Here the magnitude in that region is shown to be very small, and much less than the increase from 0 to +0.5V when Vg-k is actually driven positive.

[turbofeedus]

Hey Martin and gang,
Bit of a resurrection to show my utracer build. The kit was very straightforward and well written.
In the middle of assembling the final tester now, came across a couple things, maybe some people here could explain?

1)I've got the mains coming in on an IEC, with safety ground lifted for the moment (see #2). Wiring is simple; hot -> 1.5A fuse - > SPST switch -> power plug for laptop supply. Neutral is wired directly to power plug. (Note: I plan to change to DPDT to switch both hot and neutral, for good practice).
For some reason, I'm seeing ~2.2VAC across the two power plug terminals, even when the switch is open. I see the expected mains voltage when closed. What is the explanation for this unexpected potential in the "off" state?

2) I've hesitated to connect the safety ground, because I believe the the RS232 case is connected to circuit ground in the utracer and/or in the laptop through the USB. Will I run into any problems with the RS232 case shorting to the chassis, with or without safety ground attached?

utracer-1.JPG

utracer-2.JPG

utracer-3.JPG

utracer-4.JPG

utracer-5.JPG

[martin manning]

I'm running mine from a bench supply which is isolated, so the enclosure is not earthed. The OV side of the power supply is connected to the metal inclosure, as is the uTracer ground, and the comm jack (I'm using a USB-TTL cable, so I don't have a D9 RS232). I have a power switch in front of the uTracer board, so to power up I have to turn on the power supply and the uTracer.

[turbofeedus]

I'm running mine from a bench supply which is isolated, so the enclosure is not earthed. The OV side of the power supply is connected to the metal inclosure, as is the uTracer ground, and the comm jack (I'm using a USB-TTL cable, so I don't have a D9 RS232). I have a power switch in front of the uTracer board, so to power up I have to turn on the power supply and the uTracer.

I see, yes I did my testing with a bench supply so I could limit current as Ronald suggested, but I don't want to be lugging that thing around. It's a dual channel linear supply, so a bit of a beast. Anyway, I picked up this Dell PS for a whopping $4 from a local seller who was kind enough to let me pick it up even in these times (we did a contactless exchange).

Interestingly (or maybe super uninterestingly to those who would know what's going on), after I changed the wiring so the neutral was being switched as well, the strange 2.2VAC is gone. I don't think it was a miswiring, and I'm sure there's some explanation. Anyway I'm glad it's gone because it threw me for a loop.

I'm just not sure what to do with the earth connection. My gut tells me to connect it, and let the RS232 casing be earthed as well through the chassis, but I don't know.

utracerpowerwiringdpdt-1.jpg

[martin manning]

I'm just not sure what to do with the earth connection. My gut tells me to connect it, and let the RS232 casing be earthed as well through the chassis, but I don't know

I don't think that will be a problem. Chassis ground is uTracer ground. The things to be aware of are that the cathode terminal and one of the heater terminals are at supply voltage, and do not let the grid terminal contact ground (the grid voltage output transistors will be destroyed immediately).

[turbofeedus]

I'm just not sure what to do with the earth connection. My gut tells me to connect it, and let the RS232 casing be earthed as well through the chassis, but I don't know

I don't think that will be a problem. Chassis ground is uTracer ground. The things to be aware of are that the cathode terminal and one of the heater terminals are at supply voltage, and do not let the grid terminal contact ground (the grid voltage output transistors will be destroyed immediately).

Gotcha, good to know. Thanks Martin. I'll post another comment here when I get it finished. I did the socket wiring today, hoping to get something running tomorrow.

[turbofeedus]

Finished up the build for now, and did a few "quick tests" of an 7189. Seems to be working well!

utracer-6.JPG

utracer-7.JPG

[turbofeedus]

OOF! Popped my first fuse today!
Tried an EL34 and the fast blow 1.5A in the fuse circuit went kablooey.
I know the filament startup current can be several times the datasheet spec, so is this the right time and place for a slow blow?
Seems like an external heater supply is the first mod people go for, is this one of the reasons?

[martin manning]

I have a slow fuse in mine. The problem with the on-board heater supply is that it is HF PWM driven, and very much affected by lead inductance. See here: https://ampgarage.com/forum/viewtopic.p ... 41#p295841 Even if you remove the ferrite beads you have to compensate by setting the voltage high. Can be worked with, but IMO not worth the trouble. I use a bench supply.

[xtian]

Yes, slow blow is good for mains and heater fusing.

[turbofeedus]

I have a slow fuse in mine. The problem with the on-board heater supply is that it is HF PWM driven, and very much affected by lead inductance. See here: https://ampgarage.com/forum/viewtopic.p ... 41#p295841 Even if you remove the ferrite beads you have to compensate by setting the voltage high. Can be worked with, but IMO not worth the trouble. I use a bench supply.

Understood I'll make that change. I wonder if this is something I could use my bench supply for, to provide a baseline for the filament and then make adjustments to compensate for the onboard heater supply with the config file?

Yes, slow blow is good for mains and heater fusing.

Gotcha, will do.

[martin manning]

I have a slow fuse in mine. The problem with the on-board heater supply is that it is HF PWM driven, and very much affected by lead inductance. See here: https://ampgarage.com/forum/viewtopic.p ... 41#p295841 Even if you remove the ferrite beads you have to compensate by setting the voltage high. Can be worked with, but IMO not worth the trouble. I use a bench supply.

Understood I'll make that change. I wonder if this is something I could use my bench supply for, to provide a baseline for the filament and then make adjustments to compensate for the onboard heater supply with the config file?

Exactly. That's how I made the voltage correction curve in the link. As I said at the time, the correction will be different for each individual uTracer.

[turbofeedus]

I have a slow fuse in mine. The problem with the on-board heater supply is that it is HF PWM driven, and very much affected by lead inductance. See here: https://ampgarage.com/forum/viewtopic.p ... 41#p295841 Even if you remove the ferrite beads you have to compensate by setting the voltage high. Can be worked with, but IMO not worth the trouble. I use a bench supply.

Understood I'll make that change. I wonder if this is something I could use my bench supply for, to provide a baseline for the filament and then make adjustments to compensate for the onboard heater supply with the config file?

Exactly. That's how I made the voltage correction curve in the link. As I said at the time, the correction will be different for each individual uTracer.

I see, thank you for that link. So it seems like a 2A time delay is actually the way to go.
I think I will do this today. I'll use my bench supply to provide the 6.3VDC straight to the socket pins, and test for plate current across a couple different tubes, then use the internal heater supply and test again. I'll control for the time heated, so hopefully I get some consistent measurements.