We have one. It started to loose accuracy on a low DC volt range, but otherwise is certainly a benchmark. Every good meter deserves another one ( ... to check if either is drifting off).
Fluke must absolutely love yearly cal requirements, as so many just buy a new meter than pay the cal cost.
Fluke must absolutely love yearly cal requirements, as so many just buy a new meter than pay the cal cost.
Am I right in thinking the choke would ease the instantaneous current surge into the cap after the choke? Also would I be able to change the final B+ by varying the input capacitance size and hence changing the reactance in the choke? These might seem dumb questions to some but unlike others on this forum I don't have all the answers 😀😀
I have an old fluke 8060A which I absolutely love, however it is getting really slow to measure AC voltage ranges. Takes about 5 seconds to get up to 240V which is really annoying. At first I thought it was the sampling rate as I have a brand spanking new 179 fluke which displays almost straight away but surely there must be something else at large....😕
Yes. However, in tube amps, this surge is not a concern to the cap, only the rectifier, if it is a tube type rectifier.
Essentially, yes. The limitting cases for a choke of significant inductance are:-
a) the first cap is very large - in this case the ouput voltage (at the 2nd cap) is close to the peak of the input AC;
b) the 1st cap is not installed - in this case the output voltage corresponds to to the average value of the input AC.
Thus the two extremes with varying teh 1st cap size are 0.9 of the RMS input to 1.4 of the RMS input.
However, if the DC ouput is not close to the peak value you get with a large 1st cap, the regulation is poor - as the DC load varies from nothing to full, with a small 1st cap, you get the output DC voltage falling from 1.4 to 0.9 times the RMS input. The traditional way to fix this (if it matters - with Class A it doesn't) is to add a resistor in parallel with the output to hold it down. The voltage drop with load mostly occurs from zero to light load in practical circuits, so the resistor only has to draw a light load current.
Last edited:
I'm by no means an expert but it sounds like the microcontroller inside is running at a low clock speed for some reason... (If it has one!)
Might be time for a recapping. something ain't oscillating right.
http://www.ens-lyon.fr/DSM/AGREG-Physique/oral/Notices/N069-017.pdf
Yep, thought so, microprocessor controlled.
Though the older stuff did count rather slowly might be just a feature of the grand old designs 😛
Last edited:
TV7, I also typically use Shoog's configuration for the reasons he identifies - it nicely alleviates rectifier hash that others find they have to spend threads and threads discussing, and throw every form of high tech diode and tuned snubber at, and have to resort to separate filament transformer to avoid the coupling path in via the heater.
Regulation from 0 to 100% is certainly not great, but as Keit indicates it has no influence for class A, and for that matter for most audio content in class AB (unless you're doing continuous tone testing at high power levels). Idle level is also likely to be above 50% load already, and for instrument amps it is an essential characteristic (as is some ripple level from a few studies I've seen).
Regulation from 0 to 100% is certainly not great, but as Keit indicates it has no influence for class A, and for that matter for most audio content in class AB (unless you're doing continuous tone testing at high power levels). Idle level is also likely to be above 50% load already, and for instrument amps it is an essential characteristic (as is some ripple level from a few studies I've seen).
If it works normally on all other ranges, then it is not likely to be a fault associated with the microprocessor. This multimeter performs a basic self test when you switch it on. An extra self test is available by operating the range switches in a certain way. Try it and see what it tells you. Otherwise, it is probably some fault with or associated with the RMS-to-DC converter. I expect the fault also manifests itself by inconsistent inaccuracy on the AC range for different applied voltages.
The basic thinking is that if it is only defective on one or two ranges, then the fault is most likely be in circuitry used only in that range. For the AC range, that is the RMS-to-DC converter circuit. I cannot imagine how a fault in the range switching could give the symptom you described - a range switching fault would be more catastrophic.
Cheers for that, after reading the manual through It is amazing how much functionality this meter has, even more than the new ones which are IMO aimed at the dumb sparky in mind and cannot even differentiate between AC and DC voltages. Don't get me wrong, the T5-1000 is handy when you're checking for voltage up a ladder or chasing continuity in a switchboard, I use one for this reason.
But the 8060A passed all the tests and the calibration seems to be excellent even after all this time.😀
I'll give the ballast a go later and check the results.....
Yes to both questions. Reducing current surge is good to eliminate ringing of the transformer and will also help prolong the life of your cap. Typically you can tune your B+ by varying the first cap from about 0.47uf up to about 2uf, however its probably not the best approach. Using an extra RC stage will bring more benefits in reduced ripple.
Shoog
Yeah I was thinking it might just make it easier for the regulator and reduce wasted heat and power by tuning the input rather than another RC stage, but I can see the disadvantages also. Cheers Shoog, and others.
PSU2 models this aspect of CLC power supplies quite well, so its worth modelling to get you into the right ball park and save hassle.
Shoog
Shoog
I found this an extremely interesting thread.
I also used to work for a very small electronic ballast company and made my own ferrite E- core prototypes.
I found hiding in my basement a box of NOS GE 89G489 magnetic ballasts...simple things with just 2 wires, intended for 120 V, 60 Hz, 4-6-8 W fluorescent lamps...probably the linear tube ones, predating CFL's. '.158 A' (no leading zero...does that hint at their age?)
I never expected them to be good for anything, and inertia kept me from getting rid of them.
~ 45x52x15 mm lamination stack, 2 unsightly welds (subjective, I realize...)
I might have done an inductance measurement long ago and was unimpressed with that also.
But today I decided to visit someone with equipment that could measure the inductance with DC bias...
Started with ac only (dry, like a martini)...
61.4 ohms DC R
0.35 vrms 60 Hz 1.008 H
1 v 1.147 H
5 v 1.401 H
9.86 v 1.487 H (instrument complained about 10 v...too low Z I suppose)
Set to 9.8 vrms 0 mA DC 1.501 H
Aborted this drive level as it declined to run 15 mADC bias after accomplishing 10 mA.
Dropped down to 1 vrms 60 Hz for the rest of my experiment...
1 v 60 Hz 0 mA DC 1.189 H
1 mA DC 1.194 H
Skip some data...add later if interesting...
L 'peaked' at 5 mA DC bias 1.1945 H
Tested at 1 vrms 60 Hz, 0-200 mADC.
Points of particular curious interest to me:
At 10 mA, inductance had dropped to 1.189 H, the 'dry level'.
At ('nameplate') 160 mA (but DC), 1.0365 H (about 13% reduction in L vs. 'dry'.
At 200 mA DC, 0.980 H (17.6% reduction of L).
This was just some experimentation with arbitrary subjective choices and observations on my part.
I had not figured out how to use the DC bias in past infrequent visits, so I learned a few things today. Still not sure what I'll do with them other than experiment...
Another time we hipot tested a 70 V audio transformer up to 1000 V DC for curiosity's sake...gotta have a hobby, right? (I don't smoke & bowling just doesn't do it for me).
Maybe later I'll graph the rest of my data to see what it looks like.
I also used to work for a very small electronic ballast company and made my own ferrite E- core prototypes.
I found hiding in my basement a box of NOS GE 89G489 magnetic ballasts...simple things with just 2 wires, intended for 120 V, 60 Hz, 4-6-8 W fluorescent lamps...probably the linear tube ones, predating CFL's. '.158 A' (no leading zero...does that hint at their age?)
I never expected them to be good for anything, and inertia kept me from getting rid of them.
~ 45x52x15 mm lamination stack, 2 unsightly welds (subjective, I realize...)
I might have done an inductance measurement long ago and was unimpressed with that also.
But today I decided to visit someone with equipment that could measure the inductance with DC bias...
Started with ac only (dry, like a martini)...
61.4 ohms DC R
0.35 vrms 60 Hz 1.008 H
1 v 1.147 H
5 v 1.401 H
9.86 v 1.487 H (instrument complained about 10 v...too low Z I suppose)
Set to 9.8 vrms 0 mA DC 1.501 H
Aborted this drive level as it declined to run 15 mADC bias after accomplishing 10 mA.
Dropped down to 1 vrms 60 Hz for the rest of my experiment...
1 v 60 Hz 0 mA DC 1.189 H
1 mA DC 1.194 H
Skip some data...add later if interesting...
L 'peaked' at 5 mA DC bias 1.1945 H
Tested at 1 vrms 60 Hz, 0-200 mADC.
Points of particular curious interest to me:
At 10 mA, inductance had dropped to 1.189 H, the 'dry level'.
At ('nameplate') 160 mA (but DC), 1.0365 H (about 13% reduction in L vs. 'dry'.
At 200 mA DC, 0.980 H (17.6% reduction of L).
This was just some experimentation with arbitrary subjective choices and observations on my part.
I had not figured out how to use the DC bias in past infrequent visits, so I learned a few things today. Still not sure what I'll do with them other than experiment...
Another time we hipot tested a 70 V audio transformer up to 1000 V DC for curiosity's sake...gotta have a hobby, right? (I don't smoke & bowling just doesn't do it for me).
Maybe later I'll graph the rest of my data to see what it looks like.
Well done for realising what you had in the basement!
I have continued to use and compare the measurement technique presented in post #4 (I update the linked doc as needed and just updated it the other day with a test jig doc). What measurement instrument did you use that provided DC bias ?
Even 1H can provide a considerable ripple reduction in valve amps, especially SE and UL types where B+ ripple is not inherently neutralised, and as a way to attenuate screen and preamp stage supply ripple.
I recall one advantage of being in a 240Vac mains country is that those vintage flouro chokes need a higher inductance for a given tube.
I have continued to use and compare the measurement technique presented in post #4 (I update the linked doc as needed and just updated it the other day with a test jig doc). What measurement instrument did you use that provided DC bias ?
Even 1H can provide a considerable ripple reduction in valve amps, especially SE and UL types where B+ ripple is not inherently neutralised, and as a way to attenuate screen and preamp stage supply ripple.
I recall one advantage of being in a 240Vac mains country is that those vintage flouro chokes need a higher inductance for a given tube.
The instrument was a Wayne-Kerr...didn't note the model, but looking on Google, I think it was something like a 3255. Not a hobbyist instrument due to cost and complexity. Wait, maybe that's not true if you look at what some people pay for their audio gear. I don't have easy access to it very often...don't want to abuse the generosity...I don't have the money or space to own one...heavy, too. I did almost buy a different model on eBay once for a few hundred dollars with no manual but got scared off because it was a joint project between Wayne-Kerr and someone else like Philips. It was discontinued, and the US office of Wayne-Kerr didn't have a manual either (!!)...just a 4-page calibration verification procedure that was translated to English by a non-English speaker. I decided I'd let some other bidder win...nothing more discouraging than a bargain that has few support options, all probably involving far more money than the purchase price.
I remember a ballast 'failure analysis' I was asked to participate in. An end user installing fluorescent light fixtures with electronic ballasts in workstations they exported to another country with 240 V/50 Hz mains. The lamps were exploding and a non-technical person was hell-bent (I could tell because I heard him ranting) on blaming the ballast manufacturer. (So they ask a small local company for 'expert opinion'). The ballasts were high quality from a major international manufacturer. My boss asked me to look over the provided ballast specimen. It had spring-clamp push-in terminal blocks instead of flying leads. I said that giving me the ballast removed from the installation was far from a complete story, but what I could tell from the scratch marks around the terminal blocks was that the fixture wiring was installed into the wrong terminals, effectively bypassing the ballast. IIRC the lamp running voltage (once lit) was only a little over 100 Vrms. Apparently under the circumstances, the 240 vac was enough to ignite the lamp (or I don't remember from 1994 if it was actually a higher mains voltage)...but the important part was that in USA, the mains voltage was probably too low to ignite those lamps without the starting circuitry, and would have remained off. But the higher input voltage in the other country combined with what looked like a certain wiring error in the workstation assembly process was enough to ignite the lamps with no current limiting. I think I had a hard time keeping a straight face (the angry guy had already left) when I told my boss what my opinion was...that they needed better work instructions or inspection or both...I also thought it was essentially unused because they had bypassed the internal circuitry...I was sure it would still work.
He agreed and told the man he didn't seem to have anything to support his claim, based on what we were able to tell from the removed ballast. I was surprised how scratched the terminals were, but it made it obvious which ones had wires inserted and which did not. :O)
I imagine that was an expensive and embarrassing production run.
I remember a ballast 'failure analysis' I was asked to participate in. An end user installing fluorescent light fixtures with electronic ballasts in workstations they exported to another country with 240 V/50 Hz mains. The lamps were exploding and a non-technical person was hell-bent (I could tell because I heard him ranting) on blaming the ballast manufacturer. (So they ask a small local company for 'expert opinion'). The ballasts were high quality from a major international manufacturer. My boss asked me to look over the provided ballast specimen. It had spring-clamp push-in terminal blocks instead of flying leads. I said that giving me the ballast removed from the installation was far from a complete story, but what I could tell from the scratch marks around the terminal blocks was that the fixture wiring was installed into the wrong terminals, effectively bypassing the ballast. IIRC the lamp running voltage (once lit) was only a little over 100 Vrms. Apparently under the circumstances, the 240 vac was enough to ignite the lamp (or I don't remember from 1994 if it was actually a higher mains voltage)...but the important part was that in USA, the mains voltage was probably too low to ignite those lamps without the starting circuitry, and would have remained off. But the higher input voltage in the other country combined with what looked like a certain wiring error in the workstation assembly process was enough to ignite the lamps with no current limiting. I think I had a hard time keeping a straight face (the angry guy had already left) when I told my boss what my opinion was...that they needed better work instructions or inspection or both...I also thought it was essentially unused because they had bypassed the internal circuitry...I was sure it would still work.
He agreed and told the man he didn't seem to have anything to support his claim, based on what we were able to tell from the removed ballast. I was surprised how scratched the terminals were, but it made it obvious which ones had wires inserted and which did not. :O)
I imagine that was an expensive and embarrassing production run.
