Bottlehead Forum
Bottlehead Kits => Legacy Kit Products => Topic started by: MarimbaMan on February 21, 2011, 07:23:23 AM
-
I finally pulled the Parabees out of the closet and built them after 6 years. All resistance checks out fine, except
for T1 and T2 on both sides.
Manual spec on these terminals is 155-165 ohms
My readings:
T1 T2
Left 910k ohms 970k ohms
Right 810k ohms 910k ohms
I double checked all connections, the RRSFs are connected to 1S and 1F on the transformer. The 10 ohm resistors are
not resting on the paper cover on the transformer. The UF4007s are oriented correctly, as are the filter caps. I
double checked the bleeder resistor, it is attached properly. It appears I made the same mistake on each amp, since
the readings are so close. any ideas? thanks in advance
-
Fortunately, I found my old copy of the manual! You are measuring the DC resistance of the power transformer high-voltage winding, from each end (through the 10 ohm resistor) to the center tap which is grounded.
I would guess that the connection of the ground buss to transformer terminal 1T (the center tap) is not good.You can check for continuity between the transformer terminal and T4/T14/T16/RCA jack outer conductor. You may find that the connection is intermittent if you wiggle the wires or terminals a bit.
That ground buss magnet wire was very difficult to solder, requiring a lot of heat to break down the insulation, and there is a caution to not overheat 1T. If that joint is bad, you may want to detach the buss wire from 1T in order to tin the buss wire at high temperature, then re-attach it once the insulation has been burned off.
-
Paul,
My kit shipped with naked ground buss wire, not magnet wire. I checked for continuity from transformer terminal 1T and T4/T14/T16/RCA jack outer conductor. The connection is good between all terminals, on both left and right amps. I know my multimeter is good, because all the other resistance checks were right on. Still scratching my head....
Thanks
-
Oh, excellent - your grounds are pretty good now.
Time to trace the circuit. Check resistance from 1S to 1T, and 1T to 1F. Then move downstream, the other end of the 10-ohm resistors and the T1/T2, and down the ground buss.
-
1S --> 1T 10 ohms
1T --> 1F 0
1S --> 2T 0
1F --> 2T 10 ohms
both right and left read exactly the same. so far so good
Right Left
T1 --> T5 .691 M ohms .747 M ohms
T2 --> T5 .755 M ohms .726 M ohms
T1 -->T14 .850 M ohms .850 M ohms
T2 -->T14 .850 M ohms .850 M ohms
I checked the diodes again, silver banded side on all four face T4/T5
-
Those are startling measurements, so it's time to be very clear and methodical in order to get to the bottom of this. It's probably quite simple, once we figure it out - it usually is.
There is no tap on winding 2 of a PGP8.1 power transformer, therefor no 2T terminal. I can't tell whether you have a different transformer, made a typo when transcribing the terminal identification, or the terminology is confusing - but the latter seems quite likely, so I'll try to be boringly clear:
The power transformer should be a Magnequest PGP8.1, with solder terminals on both sides of the winding. With the chassis upside down, and the power transformer toward the back the rearmost side of the winding from left to right should be labeled 1T, 3S, 3F, 5S, 5F, 2S. Across the front, the labels from left to right should be 1S, 1F, 4S, 4F, 2F. See the manual, page 18. (Incidentally, S and F refer to the start and finish of the winding respectively.)
I have used the terminology T1, T2, T3, T4, T5, etc. to refer to the terminal strip solder lugs - see the manual, page 25. It is easy to confuse T1 with 1T, etc - for which I apologize but we are stuck with this system. Just be careful.
Winding 2 is the primary, connected to the power line and switch. The high voltage secondary is winding 1. Windings 4 and 5 are both 2.5v; they are wired in series to get 5v for the 300B. Winding 3 is the 6.3v winding for the driver filament power.
If all that is consistent with what you have, then you should be able to measure the winding resistance of the high voltage secondary. It should be around 150 ohms from 1S to 1T, 150 ohms 1T to 1F, and 300 ohms 1S to 1F. If your first two measurements use the terminology as above, then they are very wrong. This may be due to the circuitry connected to these taps, so you may have to unsolder and remove all connections to 1S and 1F to make these measurements accurately.
-
The trannie is an MQ PGP8.1
You were correct, I did confuse 1T and T1. My apologies.
I did remove the RRSF from 1S and 1F, and got the following measurements:
1T to 1F = inf
1T to 1S = inf
1F to 1S = inf
no resistance at all between these terminals.
-
The terminals on the PGP 8.1 usually had some varnish on them that could insulate the tab. Make sure you are getting good electrical contact with the DMM probes. Scraping the terminals with an Xacto knife will usually get through the varnish.
-
Yes, the varnish is honey-colored, and the terminals appear to be bare metal, but I scraped them anyway. After scraping, I got the same readings, no resistance
-
There are hair fine wires that come out of the coil and wrap around each of those terminals. It's worth looking, maybe with a magnifier to see if they have broken away from the terminals during construction.
-
Some digital meters get confused when trying to measure transformer windings. I think it is the high inductance...John
-
On the face of it, an infinite resistance would indicate the winding is broken internally. But it seems improbably that both halves of the HV winding would be broken, and even less likely that this would be the case in both power transformers. Try to measure winding 2 - around 10 ohms if I recall correctly. Let's be really sure we are making a good measurement before suspecting a bad transformer.
-
Doc -
I have a light/magnifier on an articulated arm, 1T looks like a solid connection of the hair wire, 1S and 1F appear to have the wire connected, although it is much harder to see those due to the angle. I will probably need to pull the 8.1 to be 100% sure on 1S and 1F.
I am using a Rat Shack digital mm, model 22-805, fresh battery
Paul -
I measured 2F to 2S, and got 5.3 ohms. I don't know if this is good news or bad news. I agree with you, though, that it is statistically extremely unlikely that I got two bad transformers, and that they are bad in exactly the same places.
-
Just a thought - are you using the continuity tester setting for the 1S/1T/1F measurements, or autorange ohms setting or a selector setting of around the 20K ohms range? The continuity checker only measures up to about 100 ohms on some testers and calls anything above that infinity.
Anyway I think at this point it might be best to pull the tubes, disconnect the rectifier connections so that the transformer is disconnected from the power supply rectifiers at 1S and 1F and carefully clip meter probes to 1S and 1T. Set the meter to the highest AC Volts setting (which must be at least 500VAC). Plug in the amp and turn it on and see if you get any voltage.
-
Nice meter - I have two of them! It's autoranging on volts and ohms; 750vAC/100vDC maximum. Measure voltage from 1T to either 1S or 1F; between 1S an 1F there will likely be a bit over 750 vAC.
-
Voltage measurement:
1S to 1T 334 VAC
1S to 1F 679 VAC
-
I forgot to mention, I am using autoranging on the mm
-
Other than being a bit lower than expected, those look good. Perhaps you line voltage is a bit below 120v? Anyhow they are symmetric, so the transformer is good (all the winding turns are there). Your previous resistance measurement problems must have been due to poor connections. Time to solder the 10 ohm resistors to 1S and 1F, and check resistances from each end of each resistor to 1T - i.e. confirming the solder joint. Then proceed downstream, one joint at a time.
-
The 10 ohm resistors I mangled are being shipped by Mouser. In the meantime, here are the resistance checks from the other amp (wired identically):
1F to T2 = 10.1 ohms
1S to T1 = 10.9 ohms
T1 to T5 = .740 Megaohms
T2 to T5 = .705 Megaohms
T5 to 1T = 7 Megaohms and dropping
T4 to 1T = 0
Hope these are the measurements you are looking for, otherwise, let me know.
-
Ok, the RRSF has been rebuilt with new wirewound resistors, and another fresh battery in the MM.
Left amp measurements are in the previous post, here are the new right amp measurements:
1F to T2 = 10 ohms
1S to T1 = 10.1 ohms
T1 to T5 = .668 Megaohms
T2 to T5 = .713 Megaohms
T5 to 1T = 7 Megaohms and dropping
T4 to 1T = 0
T1 to T5 and T2 to T5 are across the diodes, do these look correct?
-
Try swapping the leads on your meter when testing the rectifiers. They should read high like you are seeing in one orientation, and with the meter leads swapped end for end on the rectifier they should read lower, like maybe 1500 ohms. The only time to be concerned is if you see a really low reading like under 100 ohms in each orientation. That would mean that the rectifier is shorted out. When the rectifiers are measured in circuit the values can differ somewhat from what I describe here, but very low ohms measured across a rectifier is always bad.
-
I swapped the meter leads and remeasured:
Right amp
T1 to T5 = inf
T2 to T5 = inf
Left amp
T1 to T5 = inf
T2 to T5 = inf
-
This is problematical - it may be that the diodes are dead, or it may be that your meter does not use enough voltage for this test to be valid, or it may be that you are using the Cree SiC diodes which need more voltage to turn on.
-
I broke out my $10 analog meter and got the same readings as above across the diodes, and inf again when I swapped the meter leads. Is there a way to determine whether or not I have Cree SiC diodes?
-
The Cree diodes were a change that some people made; if you did not make this change then you don't have them. (They are TO-220s, large chips with a large metal heatsink, and inconvenient to mount - made for PC boards.)
So, the diodes are toast apparently. They were 1000v UF4007s originally (the Crees are rated 1200v, for more margin - that's why some people used them.). The UF4007s should be easy to find.
-
The diodes came with the kit, so they are not Cree SiC. I am the only one who touched them, so it appears I fried them when soldering. I'm regrouping and ordering new diodes, and will post again once the amps are put back together. Thanks all for sticking with me through this!
-
I got the replacement diodes in the mail from Queen Eileen, and did a quick measurement straight out of the box:
.718 Megaohms, reversed the leads, and got inf. It appears these are the same measurements as the old diodes.
I then installed the new diodes, and did the resistance check:
T1 = .718 megaohms
T2 = .830 megaohms
This appears to be correct since the diodes pretty much read the same uninstalled as they do installed.
At this point, I'm wondering if I should approach this from a different perspective:
The manual states that the resistance check should be:
T1 155-165 ohms
T2 155-165 ohms
How is this range calculated? Or, what components are included in these measurements from the terminal to ground?
-
Sorry, I have to ask, because I'm an Engin-nerd, what meter are you using? The meter has a lot to do with the readings you get. Some aren't so good. Some are.
New Cree diodes should measure a forward diode drop in the forward direction and a high resistance (or infinity/overload) in the reverse direction.
-
.718 Megaohms, reversed the leads, and got inf. It appears these are the same measurements as the old diodes.
This appears to be correct since the diodes pretty much read the same uninstalled as they do installed.
It's a problem that revolves around using an autoranging meter. It should read more like 500 ohms, not 718K ohms. I'm measuring about 450 ohms in the forward direction on some UF4007s out of the same bin that your replacements came from using our old reliable Radio Shack meter, and 490 ohms with an HP3466A that we keep on the bench. Both are set on diode check, not autorange. On autorange the HP reads something crazy like 256K ohms. If you have a diode check setting on your meter, use that. If not, try using the meter on a manual range setting of 2K ohms.
-
I am using a Radio Shack digital mm, model 22-805, fresh battery. I'm not using Cree diodes, I'm using the stock UF4007's.
Doc, I measured the newly installed diodes with the diode check on the mm:
T1 474mV
T2 481mV
-
Readings from different meters will vary. That is, one might show 200k ohms, one infinity, one 500k ohms and some read to over a mega ohm. The RS meter might not show diodes the same as a more expensive meter.
Yes, I'm a big fan of Fluke meters. You see one in one of the pictures of Doc's workbench somewhere on the site. They give very reliable readings and usually have a diode checker position on them. You can get a resistance in one direction and swapping the leads with my meter you get the forward voltage drop, most are under 1V, some just over, in the other direction.
-
The values look fine but I think we have a bit of an issue with the unit of measure. mV is the abbreviation for millivolts. We would be looking for 471 ohms which is equal to .471 Kohms, and the scale on the diode checker setting on the meter should indicate one of those.
-
I reread the diodes with the diode setting on the RS meter, it is indeed measuring in mV. Since every resistance reading was spot on except T1 and T2, I'm now wondering if the meter is defective measuring across diodes, or maybe wasn't built for this application.
In the interest of completing the project, I wouldn't mind spending the money on a good Fluke, although I don't want a $500 meter. Doc/Grainger, any recommendations on a good Fluke model?
-
OK, so perhaps you meter is attempting to measure the forward voltage drop across the diode. I think the UF4007 is spec'd at something like 1.7V rather than .48V so something still doesn't seem quite right with the meter reading. At any rate the diodes seem to be functional and I think you can move on to whatever else might still not be working right. Rather than buying a meter can you possibly borrow one from someone? That way you can compare its readings with the RS meter before deciding if you need to replace it.
-
1.7v is at high current, 1-2 amps. Meter test current is likely around 1mA and should drop around 0.6v according to the data sheet. 470mV seems close enough to me.
Here's a link to a data sheet:
http://www.datasheetcatalog.org/datasheet/GeneralSemiconductor/mXyzrxzq.pdf
-
I'll see if I can scare up another mm at work tomorrow.
-
MarimbaMan,
PM me and I will send you some links to reasonably priced meters.
-
I borrowed a Fluke 70 III from work and remeasured:
diode check:
.448 reversed leads = inf
.451 reversed leads = inf
the value above was not specified, I'm assuming it was ohms
resistance
T1 = 1.3 Megaohms
T2 = 1.4 Megaohms
I'm still miles away from the spec of 155-165 ohms
-
Are you clipping the black DMM lead to a solder terminal on the ground buss when you do this test? If you are clipping to the copper ground buss wire rather than a solder joint that it attaches to, you may not be getting continuity because of the varnish coating on the wire.
-
The kit shipped with naked ground buss wire, not magnet wire. I attached the negative lead to the buss near the rca tab to do all my measurements, to be on the safe side, I clipped the neg lead to the solder joint on the rca tab, and got the same readings.
-
OK that's good. What's the resistance reading from the ground buss to 1T on the power transformer?
-
ground buss to 1T = 0 ohms
-
OK, now try measuring resistance from 1S on the transformer to that same point on the ground buss, and then from 1F on the transformer to the ground buss. Use either autorange or manually set the meter to the 20K ohms range.
-
set meter to autorange:
ground buss to 1S = .820 Megaohms
ground buss to 1F = .923 Megaohms
-
That sounds like the ground that the transformer gets might be wrong. I don't have a schematic so I have no idea where the transformer/power supply ties back to the ground bus. Try heating up the grounding connections and the ones that should give you the resistance (to ground??) Doc called for. I tried looking but couldn't find the number for 1S and 1F.
-
No, 1T is the ground tap, and that's OK. The rectifier's are OK, the high voltage secondary is putting out about the right voltage, it's just these resistance readings that are weird and there doesn't seem to be any good reason for them to be that way other than the meter being fooled by something in the circuit. Without actually having eyes on the amp I can't really come up with anything else to test that would be very relevant. At this point if it was in my lab I'd probably fire it up and see what kind of voltages I get, particularly if all resistance measurements are now the same on the good amp and this one (and honestly I'm a little lost on that point since the thread has grown...)
-
To recap:
The original issue was that resistance checks on both amps were spot on except for T1 and T2 on both. This led me to believe I made the same bonehead mistake on both amps.
I installed the tubes, put a shorting plug on the rca jack, and did a voltage check on the right amp:
Terminal Spec Actual
1 365VAC 33VAC
2 365VAC 5.5VAC
15 465VDC 5mVDC
A1 75VDC 0VDC
A2 450VDC 11mVDC
A4 75VDC 0VDC
B6 187VDC 1.4VDC
B8 3VDC 11mVDC
Doc, I will be happy to provide pictures if that would help. Thank you.
-
OK, a few pages ago you had readings on the unloaded high voltage secondary (1S, 1F) that looked pretty reasonable. Now with the amp circuit reattached to the power transformer the voltages look like there is a short somewhere that is pulling things way down. Try pulling the tubes and checking the voltage at terminal 15 again. The voltage may rise higher than 465V without the tubes loading it.
-
I removed the tubes, plugged the amp back in, turned on the power, and got:
T15 9mVDC
-
I decided to do a voltage check comparison with the left amp, with the shorting plug, but without tubes.
I switched off the amp as soon as I turned it on. There is a buzz and a bright arcing at the connection of 1T and the hair-fine wire going into the transformer
-
That sounds like the transformer wire isn't continuous to the transformer lug. The high voltage would arc across the break point.
I'm just stabbing around in the dark as I know nothing about a Parabee. I suppose it is an older 300B design.
-
I decided to do a voltage check comparison with the left amp, with the shorting plug, but without tubes.
I switched off the amp as soon as I turned it on. There is a buzz and a bright arcing at the connection of 1T and the hair-fine wire going into the transformer
Look at that connection closely, preferably with a magnifier to see if the wire is disconnected or otherwise damaged.
-
I used a 10X jeweler's loupe and a Maglite, and it's still pretty hard to make out what's going on. The left amp seems to have two hair wires coming out of the transformer, and the right amp seems to only have one. Could the arcing be between the two wires? There also seems to be more varnish around the two wires on the left amp. I am very hesitant to mess with them too much, I don't want to break them or melt them.
-
Ok, I found the arcing problem, the hair wires are disconnected from the 1T tap
-
Ah hah, that would certainly account for the infinite resistance readings, assuming it was like that before the spark (which I suspect is likely as discussed back a few pages). With some care one can usually tack solder them back on. If you are successful be sure to check the T1/T2 resistance again to see if it reads correctly, before giving it power.
Once again I am sorry, I'm having a little trouble keeping track of the events. Has the other amp been given the voltage test yet?
-
To clarify, the low voltage readings above were for the right amp. I wanted to do a voltage reading on the left amp for comparison, and that's when I found the arcing problem on the left amp. I'll attempt a repair on this now
-
Sorry, you did indeed say the left amp. Definitely focus on that amp first.
-
Good news!
Left amp: I tack soldered the hair wire, and did resistance readings from 1T to T4, 0 ohms.
resistance on T1 = 157 ohms
resistance on T2 = 167 ohms
all resistance on the left amp is now within spec.
Bad news!
I've blown two 2A fuses when powering on to do the voltage check
-
Are the fuses fast blow or slow blow?
-
They are fast blow
-
Try 2A slo blos. They handle the startup surge a lot better. If they blow we still have problem to sort.
-
The 2A slo blo took 4 seconds to blow rather than instantly