Hi all

Had a rather embarrassing fault happen the last few weeks. It was with my Astor PP (1953) that I use almost every day. I'll get to the fault in a bit, a bit about the radio gram as it is a very impressive unit with 2 x 15" and 1 x 5" Speakers. It has 2 hinge units. the left one containing the Radio/Amp and the right one containing the record player and a small tube FM unit I built.

The valve line up is rather impressive as well, 6AN7, 6U7G, 6AV6, 6U7G, 6U7G, 2 x 6AQ5 (Push Pull) and 5Y3GT rectifier. This unit has been fully restored (Caps, Few Resistors, Wire due to rubber insulation etc.)

Now to the fault. I was running the radio as usual on the AM tuner tuned to Cruse 1323 (Adelaide), This station is starting to get poor in reception due to many different factors, too many to list here. The sound started buzzing a bit a first but since I was in another room I just put it down to interference. It went away and didn't return for a minute or so.

Then it returned, WOW! did it return. it buzzed and cracked and made a hum like an overloaded transformer!
Running into the room to try and turn it off before flames I could smell the transformer. Bugger! It tripped the breaker for that circuit in the house.

So, I pulled the plug anyway and went to turn back on the power at the switchboard. When I got back in I started investigating the unit. The transformer was still hot, so initial thoughts I had cooked the transformer. I put the unit back to the wall and left it there. Embarrassed that my favourite radio was dead, how could this happen to ME, I repair these things!

So yesterday I decided to pull out the chassis and see how bad the transformer was. I grabbed my trusty meter to find it flat. Bugger, I then grabbed my el-cheapo Meter. This showed me that the winding was intact in the transformer and about the resistance I would expect.
So over to the megger and test the insulation. Passed with no hint of an error. What went wrong?

So out came the valves and switched el-Cheapo over to AC volts. this gave me interesting results on the HT winding's. It was varying between 30 and 80v. (Supposed to be 255v a side). Bugger. So, I tested the 5v and 6.3v winding's. these gave me different readings as well. Now I could see that the 6.3 winding was working because I had left the dial lamps in place. These were glowing at full brightness and not flickering, so thinking the transformer is stuffed and that the bulbs are not reacting to fast changing voltage that much.

Slightly annoyed because things were not making sense, I sat in my chair for a bit looking at the bench. I then grabbed the meter and stuck it into a test lead for GPO’s that I have. I turned it on and nothing, about 30 or so volts. Bugger.

It was time to fire up the old faithful a Fluke 77 with a new battery. This time testing the transformer I got the readings I expected. With that el-cheapo meter now implanted into the rubbish bin I started looking at what caused all the fuss. There was no “short” on any of the voltage lines. So, in went the rectifier and turn on. Nothing, Not even a glow from the heater.

I started tapping the valve thinking, then with a bang, buzz and sparks inside the valve. I pulled it out and looked. There was something wobbling on the top mica. This was ether part of the heater or the anode itself. I am not sure but it was not doing what it was supposed to.

A new 5Y3GT and all voltages were back. Interesting note about the 5Y3s that I have is some have a Halo on the top of them and the one that failed didn’t. I have put one in with the halo this time to see if it is something about it running on its side as I know some valves do not like that.

To sum up, the fault may not be exactly what you think. I thought I had burnt out the transformer because it was hot and tripped a breaker and the voltages were not correct. Turns out I had 2 completely different faults. A Stuffed rectifier and a stuffed meter.

Always check your test equipment is working before using! You can save a lot of fault-finding time!

I think everyone has at some stage been led up the garden path by dubious test equipment.
It is also fairly unusual to burn out power transformers, if caught early in their overheating.
JJ

It might be worthwhile putting some fuses on the secondaries of the transformer because it might not survive another short circuit fault.

Hi

I don't intend putting fuses in the secondary of the transformer. I have never put fuses in before, I will think about it but don't want to modify the chassis.

Some radio designers put sacrificial resistors or "fusistors" in the B+.

The 6X5 has a reputation for taking out secondaries. The cathode in some versions of it progressively expands up the centre and folds over to short onto a plate.

Yes, I'm only a beginner at repairing TVs and I got caught out with a $50 Jaycar meter that stated its max DC voltage was 1000 volts, but in fact it went into OL far lower than that ,,,and my inexperienced thought it meant open circuit! !!!!!!!!,, lucky, although I'm not a rocket scientist, I am scared of electricity and I could smell a rat with that meter,,so I borrowed a meter and sure enough there was 600 volts in that power cap line,,, but Yes the meter was wrong being rated at 1000V, but a experienced person would of seen this straight away ,,, So for all beginners buy the best meter you can as you wont see the mistakes its making like a guy whos been doing it for 20 years,,,,,anyway its the bloody bin ,,more bloody Chinese land fill..


All the best ,, pete

As far as I am concerned, unless I am testing torch batteries, I will only use a Fluke meter with true R.M.S.

Wayne.

The bottom line is that you get what you pay for. There's a few good brands of digital multimeter but yes, Fluke is the best on the market and if you have the money for one - get one. Before my original Model 73 was stolen by an employee I sacked about five years ago I'd used the same meter for a good twenty years with no problems. I had it calibrated once but was told I could have gone without it. Fluke products are that good. I replaced that meter with another Fluke and I also have a Fluke tong tester.

That said, do replace multimeter batteries yearly or you'll end up cleaning up acid leaks and/or getting incorrect readings.

As for vintage radios, I've learned a few times never to run them without direct supervision. At the end of the day they are an old electrical appliance and faults in HV circuits can develop with the snap of a finger. Back when I was serving my apprenticeship with the NSW Dept. Health I used to play a Kriesler 11-20 to put me to sleep each night. One night it almost did put me to sleep - permanently. It overheated, the transformer cooked itself and filled the room with acrid smoke. I woke up just in time. In more recent times I used an AWA Model 500M as a daily listener. It went fine for around six years in my office and one day, the same thing - the transformer overheated and almost took the rest of the radio with it. I was nearby, luckily enough. The Kriesler is long gone, stripped for spares but the AWA remains on the repair line, awaiting its owner to do a heart transplant. It will one day sing again.

I quite often take swipes at test equipment, such as signal generators and calibration, Digital Voltmeters with dirty volts & RF. It is seriously important to keep an eye on all of the meters. Put them together & singularly, measure the same source, like a radio "B" voltage that will not be affected by the loading of the analogue.

Much of the stuff like my ageing VCT Valve and circuit tester & others not mine, an AVO meter, Sanwa & Peak 200H, Philco Frequency meter; have all required aging capacitors & a few resistors & in some cases wiring. The Heathkit OS-1 Oscillocope was one of the worst & its problems were solved by replacing every resistor & cap in it: All went bad. I did up the voltage rating of it's input caps to 2kV

I know of two Jaycar QM-1010 (analogue) meters where there is a significant reading discrepancy between 250 &1000 DC Range. That has to be factory and its not a loading issue. "Loading" is something to be considered when comparing voltages quoted from an analogue meter to those read with a digital. Used incorrectly and without understanding, both can lie.

What needs also to be considered is that you need to select the meter for end use, over price. Many DVM's do not like RF and cannot handle RF & DC simultaneously and some have been known to flash over internally, when presented with RF. Many bits of "Modern" test equipment are incapable of withstanding valve radio voltages. This is why I often say not to use one to measure across the outside of a transformer (pins 4 & 6 5Y3) unless it can handle over 1kV AC.

If you have issues with measuring DC on valves where RF is present. Old Analogue meters will not see the RF.

With a RCA filament in hollow tube pattern valve, there is only one position where it can be run sideways. Failure to comply will often see a tired sagging filament hit the plate with spectacular & fatal results, for the valve. The Russians (Sovtek) make a ruggedised version that looks more like a 6X4 etc. with a massive high thermal inertia single filament more like a pencil. I suspect a cathode sleeve also. It heats like a heater tube & you do not get the voltage surge.

If you copy the 6X5 with resistors in the plate circuit, they were not principally for fuses: However, they can act as "fusistors" but current squared by resistance applies. Too big a wattage & the resistor will not cook. This is also why I frown upon increasing the wattage of a "back bias resistor". Ancestors knew what they were doing: Better it burn than the transformer.

Marc.

This is an interesting topic. About 2 weeks ago I had a similar 'misleading' experience with a meter. I'm restoring a Sachsenwerk radio from around 1938, and after much work was ready to do some preliminary tests. So with all valves out I wanted to test the transformer. I connected my usually trustworthy analogue meter (20 year old Dick Smith Q-1024) to the secondaries. 30 volts, but the dial light was glowing brightly!?

After sitting back for 30 seconds or so scratching my head I melt something getting hot. A quick switch off and feeling the body of the transformer revealed that it was indeed very hot. Bugger - all this work (stripped and repainted chassis etc etc) and the impossible to find transformer is burnt out. So I was considering a rewind. I put the set to the side to work on something else, when I discovered that the AC range of the meter was not working. I still hadn't connected this with the Sachsenwerk 'problem'. Removing the back from the meter revealed a burnt section of PCB that was carbonised and shorting out the AC range.

It then dawned on me that the meter had loaded the Sachsenwerk transformer. I just hoped the meter hadn't damaged the transformer, which after testing again (this time with my Fluke 73), was fine. I was able to repair the analogue meter, but I will not use it on the high voltage AC range - just in case.

Robert

One of the things mentioned elsewhere is that Victorians are hoarders and I have hoarded, or acquired, manuals for just about every machine on the place as well as the test equipment. Much of the old stuff has not got its voltage limits on it whereas a lot of new stuff has. The second thing with newer meters is that they have a "Class" and that identifies what area they are for working in.

The oldest functional instrument here(with original box) I would like to find the paperwork for, is a Circa 1914, L.S. Starrett "Transit Level"

The major concern in valve radio, as above, is the capability of handling RF & High Voltage. What can happen is that you get the TIG welder effect, where the RF carries the DC / AC across with it and flashes over, normally with catastrophic results.

As I suggested you need to "know" the meters capability, lest you by a "Pig in a Poke", or destroy it.

Marc

I agree with the comments from members!
I have been in a privileged position where in the last 40 years I have tested/used many many multimeters
.
When you pull a Fluke or similar apart you will note the probe leads go through that standard spaced apart socket arrangement on the front panel (IP rated) and THEN IMMEDIATELY into a barrage of protection things like arc barriers, spark gaps, VDRs, current limiters, fuses before proceeding into the meter STILL MAINTAINING the separation before going into range resistors and stuff. Hence you could stupidly hook your Fluke or similar onto 240 volt while switched to the ohms range ( NOT RECOMMENDED!) and simply get an "error" reading.


By the way,that was one of my favourite "customer" test procedures for Asian meters! (That was after a uni lecturer attempted to demonstrate to his budding engineers how to measure the IMPEDANCE of the local mains by using the ohms range, what could possibly go wrong??? but that's another story.

Back to Fluke type meters, the problem was they are far too expensive for your Dick Smith/Jaycar/Altronic customer so back in the late 1990's Asian companies started to make look alikes and produced some really nice meters, I remember names like Metex , Yokogawa, Mastech that came with handbooks and calibration certificates and they maintained a similar safety standard.

BUT THEN the push and shove came to sell really cheap meters and the rot set in. I would pull a Chinese meter apart for inspection and regardless of Class / Safety ratings noted the manufactuers got rid of that unnecessary junk like safety gaps, VDR's etc etc, who needs those? and the input posts were now mounted on a PCB sometimes with a safety gap or VDR but mostly not and then the PCB tracks narrow down to the range switch and can wind up 0.5mm apart!!!!!!!!!!!!!!!. Put that baby onto a 415 volt busbar or your plate to plate 80 rectifier at high potential and guess what??? FIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIZZZZZZZZZZZZZZZZZZZZZZZZZZZZ!.....BOOOOM! I would do voltage testing on those meters (from current and so power limited sources) and just light the meters up literally ( hey look the meters got a dial light..poooof, awww its gone out!) as the arcing started up in the range switches.

So you see by about year 2000 I would have nothing to do with certifying your Dick Smith or Jaycar or Altronic meters as they were all crap in my eyes safety wise and could not comply with class or safety no matter what.

My suggestion is this: pull your meter apart and note where the input sockets go to. IF they proceed with the same spacing and then go to components that look like spark gaps, fuses, VDR's, limiting resistor and have physical barriers and/or slots in the PCB material between the tracks BEFORE going to range resisitors, switches then the class and safety rating of the meter is probably correct.
IF the input posts go straight onto the PCB then disappear straight into the range switch WATCH OUT. You cannot safely hook such a meter to a HIGH ENERGY supply like a 415 volt bus bar in a switchboard. If the meter fails when hooked to your 80 plates then the fault energy is limited but a curl of smoke may come out of the meter just before the curl of smoke from the transformer as members have found out!! Such meters are ok for transistor radio and amp use where you rarely get over 200 volts dc and that's why they get away with it.

Being out of the game now I don't know what you get from Jaycar or Altronics in 2017 but I bet they don't sell Fluke or Yokogawa!

Cheers, Fred.

stupidly hook your Fluke or similar onto 240 volt while switched to the ohms range

I recall a lecturer at tech college who did that with would be competitors of the then standard AVO and Taylor meters. This was his test of it being "student proof".

My suggestion is this: pull your meter apart and note where the input sockets go to.

Dave Jones has done a few meter reviews, such as this one: https://www.youtube.com/watch?v=gh1n_ELmpFI

He makes two good points.

The first being that he mentioned the same thing I did in an earlier post - one gets what they pay for. At the end of the day this pretty much applies to any consumer product.

The second relates to auto v's manual range selection. On a Fluke you get AC Volts, DC Volts, Amps, Ohms and Continuity functions. On the higher-end models you get capacitance and the ability to test transistors, SCRs and TRIACs. You don't get to select the range setting because it is not needed. The resolution of the digital display will automatically give you two decimal places for low voltage readings and use whole numbers for high voltage readings. I remember back in the 1980s we had all the electronics shops of the time showing a range of multimeters in their catalogues and spruiking a high number of range settings as a noteworthy feature rather than the pain in the arse that Dave Jones rightly calls it. On an analogue meter there's probably no way around manual ranging but on a digital meter it is simply not necessary. It's a case of the old adage "let the machine do the work" applying.

One other thing should be noted with both analogue and digital meters - one should know the limitations of the meter they are using. Cheap meters only measure up to around 600 volts. This is fine providing you don't place your probes across the outer terminals of the HT secondary winding on the average vintage radio transformer. You can often find 700 to 750 volts here so it is worth getting a meter that can measure 1000 volts, AC and DC.

I would also advise not skimping on the quality of probes either. Nice thick moulded probes (as Fluke supplies with its meters) are always going to be better than the hard ABS plastic ones where the tips can be unscrewed, exposing the soldered joint. The wiring to probes is single-insulated so do not expect it to be rated at 1000 volts unless this claim is stamped on the probes, again, as Fluke does with its probes.

As a tail end to all this I was once asked to inspect and supply a report on a multimeter failure as was my job as the resident old guy.
I usually only came into these matters after the normal commercial avenues had been exhausted and they needed the advice of the old guy.. The meter came to management with a stiff solicitors letter of demand and a threat of impending legal action etc etc.

Upon dismantling the meter it was obvious that the probe leads had been plugged into the 10 amp current measuring sockets as the tracks from the sockets to the internals were vaporized along with associated rectifers and the useless low VA rated fuse.
Arcing had occurred right into the gubbins of the meter and the whole thing was burnt up to the LCD display.

I traced the sale of the meter back to the retailer and unfortunately they had not followed procedure and told the customer to get lost when he presented the meter for replacement in the first instance rather than smooth things over and just replace or send on for expert comment. that only escalated the issue hence the letter and threats and claims of damage, subsequent damage, mental anguish and so on (oh yes! this is one of those professional litigants who sue at the drop of a hat)

The gist of it all was the customer claimed that the meter was not to some standard as it blew up when measuring the output of a mains wall socket clearly marked "10 amps". By the customers logic if he plugged the 10 amp meter into a 10 amp socket then the meter should read 10 amps!

Just stop for a moment and think about how you are going to answer that kind of logic!
The whole thing relies on a basic understanding of electrical quantitys that most people never get.
Yes, I did speak to this chap and his point was that if you could switch the meter to the 300 volt range and safely read 240 volts why cant you switch to the 10 amp range and safely read 10 amps? I thanked him very much and let him know we would be back to him in due course.

Now this is all serious stuff as you can wind up in court defending yourself against this kind of logic ( I have) and unfortunately you are amongst solicitors and lawyers and magistrates who also have no concept of physics or electrics and need detailed arguments of logic and points of law to make a ruling. Reality and common sense seem to drift away into blue sky.

In this case reality and common sense did not win the case, what did was the HANDBOOK supplied with excruciating detail on how to use the meter and had a paragraph along the lines of "never connect this meter across the poles of any mains power supply when plugged and switched to the 10 amp range" only they said it much better than that. When our legal people saw it, that was something they could understand and hang the customer on. Good old handbook. Now you know why handbooks have 4 pages of non-sense right in the beginning about things that only an idiot would attempt so you cover your backside against the litigants. CASE CLOSED!

Sorry Brad I got a bit off topic there................