Hi, I want to measure the current before an HV regulator. The regulator's V+ input is fed from the plus lead of a smoothing cap located after a diode bridge. The HV reg gnd input is fed from the neg lead of this smoothing cap.
Where do I place the 1 ohm resistor to measure current before the reg? In line with the V+ rail (resistor lead to plus lead of cap with the other R lead to reg's V+ input? Or is it across the plus and neg leads of the smoothing cap? Oh, and how do I size the wattage of the resistor? Under no load, the HV reg measures 178V on input, 70V on output and under load, 138V on input and 50V on output with about 28-33mA running through it. Thanks!
Where do I place the 1 ohm resistor to measure current before the reg? In line with the V+ rail (resistor lead to plus lead of cap with the other R lead to reg's V+ input? Or is it across the plus and neg leads of the smoothing cap? Oh, and how do I size the wattage of the resistor? Under no load, the HV reg measures 178V on input, 70V on output and under load, 138V on input and 50V on output with about 28-33mA running through it. Thanks!
This is a job for Ohm's law!
E = I x R
The 1 ohm goes in series with the "load". If you want to see the total draw of the regulator plus any load (if a load is applied) then it goes in series with the B+ lead before the regulator.
If you only want to see the draw of the thing being run by the regulator, then you put the 1 ohm between the regulator and the load (thing being run by it).
We then measure the voltage across the 1 ohm resistor and apply ohms law to find the current: E/R = I
How large a resistor? P = I x E
You can substitute there like for I... P = E/R x E or P = E^2/R for example.
Since you stated the max current draw and the voltage you can just pop them into the power equation and then pick a resistor wattage that is above that... higher power is always safer.
ok?
_-_-bear
E = I x R
The 1 ohm goes in series with the "load". If you want to see the total draw of the regulator plus any load (if a load is applied) then it goes in series with the B+ lead before the regulator.
If you only want to see the draw of the thing being run by the regulator, then you put the 1 ohm between the regulator and the load (thing being run by it).
We then measure the voltage across the 1 ohm resistor and apply ohms law to find the current: E/R = I
How large a resistor? P = I x E
You can substitute there like for I... P = E/R x E or P = E^2/R for example.
Since you stated the max current draw and the voltage you can just pop them into the power equation and then pick a resistor wattage that is above that... higher power is always safer.
ok?
_-_-bear
measuring HV - take care
yes, it's ohm's law.
30 to 50mA through 1r0 does not need a power resistor.
50mA is just 2.5mW and can be passed by any 0.125W or greater resistor.
But if you want accurate results you must measure the test resistor accurately.
5% tolerance is not much good if you need 0.5% in your readings.
50mA passing 1r0 will show 50.0mV +-0.1mV if that is the accuracy of the resistor and multimeter combined.
A 1% multimeter on the 200.0mV scale can read no better than +-2mV.
A 5% resistor can read no better than +-2.5mV on 50mA.
The total tolerance of these combined values will be 4.5mV.
Your reading could be anywhere between 45.5mV and 54.5mV for an actual 50mA current.
yes, it's ohm's law.
30 to 50mA through 1r0 does not need a power resistor.
50mA is just 2.5mW and can be passed by any 0.125W or greater resistor.
But if you want accurate results you must measure the test resistor accurately.
5% tolerance is not much good if you need 0.5% in your readings.
50mA passing 1r0 will show 50.0mV +-0.1mV if that is the accuracy of the resistor and multimeter combined.
A 1% multimeter on the 200.0mV scale can read no better than +-2mV.
A 5% resistor can read no better than +-2.5mV on 50mA.
The total tolerance of these combined values will be 4.5mV.
Your reading could be anywhere between 45.5mV and 54.5mV for an actual 50mA current.
thanks guys!
One elementary clarification. by "in series" you mean that the resistor should be "in line" with the V+ rail? I don't need a 2nd 1ohm resistor in the gnd rail, just the plus rail? Again, the plus and neg leads of the smoothing cap (located after a diode bridge) is feeding the HV reg board. So I just solder one lead of the resistor to the plus lead of this cap and the other lead of the resistor to V+ input on the HV board?
One elementary clarification. by "in series" you mean that the resistor should be "in line" with the V+ rail? I don't need a 2nd 1ohm resistor in the gnd rail, just the plus rail? Again, the plus and neg leads of the smoothing cap (located after a diode bridge) is feeding the HV reg board. So I just solder one lead of the resistor to the plus lead of this cap and the other lead of the resistor to V+ input on the HV board?
yes, "in series" means the output current passes through the series resistor.riotubes said:
But I have to ask: are you confident you have the skills/knowledge to work with HV and tube/valve gear?
Thanks Andrew for the clarification. I'm definitely learning as you can tell. The way I keep myself out of trouble is that I *always* ask before doing anything, and I make sure to clarify. Several times if necessary. I always use someone elses circuit that is completely spec'd. I'm comfortable measuring voltage, but haven't measured current except by using ohm's law. Thanks again.
In my parts bin, I found a 1 ohm 5% resistor and a 10 ohm 1% R. Would the latter be better to measure the current? The DVM is a fluke 87 model.
10r 1% will let you measure to 500mV +-5mV
If the fluke is 0.2% then on 2000mV scale is +-4mV.
Total error range +-9mV, for a 500mV reading is far better than +-4.5mV for a 50.0mV reading, more than ten times better.
Do the numbers, rather than guess.
If the fluke is 0.2% then on 2000mV scale is +-4mV.
Total error range +-9mV, for a 500mV reading is far better than +-4.5mV for a 50.0mV reading, more than ten times better.
Do the numbers, rather than guess.
Btw, not sure what ur doing with the regulator(s), but dropping 100volts in the regulator seems a bit excessive... you will have a relatively hot regulator there... of course if it is a low current deal anyhow, maybe the heat won't be too much.
_-_-bear
_-_-bear
It's a discrete reg using 500V mosfets and one gets a pretty big heatsink. The spec'd circuit should only be dropping 50ish volts in the reg. I'm using a 115V triad n68x isolation transformer. This goes to a schottky bridge and then a 1uF film cap with 100k bleeder strapped across it. Then this feeds the HV reg which feeds a discrete analog stage of dac. The HV reg is a Swenson design using Pimm self-bias boards and the output stage uses the same boards. The small cap, although unconventional, is spec'd to reduce ringing since the reg circuit does the smoothing anyway. I just can't seem to get it to work. I replaced the analog state with a proxy load (a 2k ohm power resistor) and the voltage still sagged to 50V when it should be 70V. It measures 70V under no load. I've built multiple HV regs and can't get it to work. Trying to isolate the problem...the transformer or the reg.
I'll insert the 10ohm 1% resistor b4 the reg and take measurements this week.
ok...... went back and looked through previous threads......... swenson regulator........ current source and series mosfets, not much in the way of error correction. that's why it bogs down, there's no effective error correction and no stable reference voltage. it's just a fancy resistor. you need some error correction and a reference voltage to get a stable output voltage. i have some ideas here, but it'll take me a bit of figuring....... hmm.... a mosfet diff amp and a zener.......let me draw something up.......
Thanks UncleJed!
The schematic for this series pass reg is linked below. The HV reg is on the right; I'm not building the preamp shown on the left. Q5 functions as an error amplifier. John spec'd a few parts for my implementation: C3=1uF, R3=180K, R7=51K.
http://johnswenson1.home.comcast.net/~johnswenson1/stereo/BDTpreamp.gif
The description of how it works is here:
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=141938&highlight=bdt+preamp
Since this is based on the self bias CCS board designed by Gary Pimm, Gary took interest and added additional circuitry here (which I have not tried)
http://www.pacifier.com/~gpimm/regulators.htm
The small size of this reg is appealing but given I'm not having much luck I'm open to building a different 70V reg if anyone has ideas....
Here's a PDF attachment showing 3 fotos of this analog stage
The schematic for this series pass reg is linked below. The HV reg is on the right; I'm not building the preamp shown on the left. Q5 functions as an error amplifier. John spec'd a few parts for my implementation: C3=1uF, R3=180K, R7=51K.
http://johnswenson1.home.comcast.net/~johnswenson1/stereo/BDTpreamp.gif
The description of how it works is here:
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=141938&highlight=bdt+preamp
Since this is based on the self bias CCS board designed by Gary Pimm, Gary took interest and added additional circuitry here (which I have not tried)
http://www.pacifier.com/~gpimm/regulators.htm
The small size of this reg is appealing but given I'm not having much luck I'm open to building a different 70V reg if anyone has ideas....
Here's a PDF attachment showing 3 fotos of this analog stage
ok, i worked something out for you that might work.... you will need a transistor rated at the input voltage x 1.5 (safety factor) as well as mosfets with the same voltage rating. the circuit as-is puts out 71V (not bad for just taking a whack at it?). the actual regulator output voltage is adjusted by the ratio of R7 and R8. the zener is the primaty reference voltage, and R7 and R8 allow you to adjust the reference at the noninverting input. the output is compared to the reference and the feedback can be adjusted with the R9/R10 ratio at the inverting input. if this looks similar to an op amp, you're right...... it's a very large single supply op amp. this isjust a basic draft, but you should get the idea. any changes at the inverting input are corrected just like with an op amp
oops..... forgot to get the schematic.......
still getting used to linux......
i used LTSpice's default transistors because i don't have any 300V bipolars in the library.... if you're using a 110V source voltage, you can go with 150 or 160V devices.
actually as shown, 2k ohms is where the regulator just begins to lose control. at 3k the voltage is 72V. at 2k it's 71.4V. after a bit of checking it's actually because of the 100ma limitation i put on the circuit by using Rser=3000 in the voltage source
still getting used to linux......
i used LTSpice's default transistors because i don't have any 300V bipolars in the library.... if you're using a 110V source voltage, you can go with 150 or 160V devices.
actually as shown, 2k ohms is where the regulator just begins to lose control. at 3k the voltage is 72V. at 2k it's 71.4V. after a bit of checking it's actually because of the 100ma limitation i put on the circuit by using Rser=3000 in the voltage source
Attachments
Thanks very much UncleJed for taking the time to develop a reg schematic for my project. Much appreciated. Can I request some add'l assistance in spec'ing suitable parts for the:
- PNP (I have BC560 on hand, but don't think that's adequate)
- LED
- MOSFETs
I have N-channel irfi820s on hand, will those work for the NMOS? They are rated 500V
I found the zeners at Digikey so will try to order from there.
Finally, can I use 1/8W metal oxides for the resistors?
Thanks again.
- PNP (I have BC560 on hand, but don't think that's adequate)
- LED
- MOSFETs
I have N-channel irfi820s on hand, will those work for the NMOS? They are rated 500V
I found the zeners at Digikey so will try to order from there.
Finally, can I use 1/8W metal oxides for the resistors?
Thanks again.
led?
what's the voltage feeding the regulator? i was having trouble finding a suitable pnp, but a pnp video output transistor for a TV or computer monitor might do the job.... i found a page of video output transistors from a toshiba data book... try some of the 2SA series on this page there are 300 and 400V types listed
what's the voltage feeding the regulator? i was having trouble finding a suitable pnp, but a pnp video output transistor for a TV or computer monitor might do the job.... i found a page of video output transistors from a toshiba data book... try some of the 2SA series on this page there are 300 and 400V types listed
Attachments
As you may recall, I'm trying to identify my error in building the analog stage of my dac...whether the problem lies with the traffo or the HV discrete regulator. Before moving onto UncleJed's HV circuit, I'd very much appreciate comments on these current (mA) calculations derived from mV measurements across a 10R 1% resistor inserted in series before and after the existing HV regulator.
Here are my measurements in order to derive current *before* the 70V regulator.
In inserted a 10R 1% resistor in series with the V+. The resistor is soldered to the plus lead of the final smoothing cap of the raw supply and the other lead soldered to the wire that goes to the 70V Reg's V+ input.
With DVM dial on mV, I took two separate measurements:
311.2 mV
310.0 mV
With DVM dial to Voltage, the measurement on the reg side of the 10R resistor is 142.8 V and 143.1 V on the other side.
So....
310.0 mV divided by 10 ohm = 31 mA. Is this correct?
Next, I removed the 10R resistor from this location, and inserted *AFTER* the 70V reg in series with the Load (a dummy load that is a 2K ohm power resistor). Here, I measured 292.5 mV across the 10R.
Again, I'm trying to troubleshoot why this reg is not outputing 70V when under load...and whether it could be the traffo or the HV reg. I'm only measuring in the 57-58V range when under load, but can easily dial in 70.0V with no load.
So 31mA going into the reg, and 29.3mA coming out of the reg. Does this eliminate the traffo as the culprit?
Context: Using Fluke 87 DVM with mV range of 400.0mV; 0.1mV resolution; accuracy = +/-( 0.1% +1)
Here are my measurements in order to derive current *before* the 70V regulator.
In inserted a 10R 1% resistor in series with the V+. The resistor is soldered to the plus lead of the final smoothing cap of the raw supply and the other lead soldered to the wire that goes to the 70V Reg's V+ input.
With DVM dial on mV, I took two separate measurements:
311.2 mV
310.0 mV
With DVM dial to Voltage, the measurement on the reg side of the 10R resistor is 142.8 V and 143.1 V on the other side.
So....
310.0 mV divided by 10 ohm = 31 mA. Is this correct?
Next, I removed the 10R resistor from this location, and inserted *AFTER* the 70V reg in series with the Load (a dummy load that is a 2K ohm power resistor). Here, I measured 292.5 mV across the 10R.
Again, I'm trying to troubleshoot why this reg is not outputing 70V when under load...and whether it could be the traffo or the HV reg. I'm only measuring in the 57-58V range when under load, but can easily dial in 70.0V with no load.
So 31mA going into the reg, and 29.3mA coming out of the reg. Does this eliminate the traffo as the culprit?
Context: Using Fluke 87 DVM with mV range of 400.0mV; 0.1mV resolution; accuracy = +/-( 0.1% +1)
- Status
- Not open for further replies.