| efflux |
Hi this is my first post here. I'm hoping someone can help.
I recently got hold of a Michaelson Audio Odysseus valve amp. As far as I know this is a class A ultra linear amp. It has 8 EL34 power tubes, 2 ECC83 and 2 ECC85 pre-amp tubes. It outputs around 45 watts.
This is a photo of one:
http://www.hifiengine.com/phpBB2/al...bd16f31f8491d5f
The original owner had not experienced problems.
When I received the amp it worked and very nicely so I am determined to get it working again.
I have a little experience of fixing my Fender Twin guitar amp which I completely rebuilt (it was also an ultra linear - not so common for Fender Twins). However I have forgotten some know how since then and this Odysseus amp is obviously a little more complicated.
This is what went wrong and what I have done so far:
The right channel of the amp stopped working and fused. There are fuses connected to the screen grid circuits. I looked inside the amp and immediately saw some problems. The main problem being burning around the high watt wirewound resistors at the filter caps. Bear with me because I have forgotten some knowledge about what everything does. You will understand what resistors I am talking about here - whatever they are called. These resistors read 23.5 ohms on my meter (so bad a state that reading the colour code was impossible). They are not the same type as appear in the original Odysseus. They look too small and low wattage and were almost touching the PCB. The right channel had such a bad connection here, due to heat, it was loose. I sorted these by replacing with suitable value high wattage wirewounds (I actually stuck two 47 ohm in parallel here - which I presume should be OK) and I checked all connections, PCB tracks etc. No other obvious problems in this respect. I also noticed that the wirewounds on the power tube cathodes were in terrible physical state. These were 470 ohm wirewound. They were only marginally off but again not the same resistor types as was obviously on the original amp. Possibly too low wattage again. Originally, it had the standard oblong shaped ceramics here. I replaced these resistors. I also discovered a completely blown screen grid resistor. Once again these resistors don't look the same as the originals. Possibly too low wattage yet again but I have left them, only replacing the dodgy one for the time being.
As you will now deduce, work has been previously done to this amp. What I know of is that the Filter caps were replaced along with 100uf caps on each power tube cathode and two 220uf caps in the preamp stage. This is documented as being done two years ago. I have not yet tested all the caps because this amp is hell to work on. The replaced caps are all correct spec as far as I know. Some resistors were also replaced but not the ones I replaced. All resistors seem OK. I've tested them all.
Next stage was turning the amp on again. I make the assumption an EL34 valve had gone because it's screen grid resistor immediately went up in smoke and the amp fused again. I replaced this tube (I was provided with new spares of the same Harma valves when I bought the amp) and the amp then turned on OK except I get serious motorboating. I even returned the original resistors (the ones that were working but not entirely satisfactory as far as I can tell) to return the amp to how it was when it worked but to no avail.
My questions start with what causes motorboating and where should I go from here? Testing all the caps may seem like an obvious start but unfortunately that is going to be physically hell on this amp due to it's design. What caps, if any, might be causing my problem? It it the filter caps? Did the blown valve or loose connected resistor damage anything else? My knowledge is not that good but it's more a question of brushing up what I once knew so I can provide reasonably detailed info.
If anybody can help me get this amp working it wold be greatly appreciated. I can supply photos if that is of any use.
Thanks a lot. |
|
|
| EC8010 |
| Welcome to the forum - shame about the amplifier. I once had a Michaelson & Austin TVA10. If you so much as looked at it in a funny way (or removed its input or loudspeaker load) it would oscillate violently at RF and the output valve anodes would immediately glow red. I suspect you may have the same problem here. RF oscillation causes heavy power supply currents and that can often provoke motorboating. The motorboating may be because a different value was fitted when those HT capacitors were replaced. The amplifier clearly has a serious fault that has been there for some time and is going to need an oscilloscope and a fair degree of intelligence to fix. |
|
|
| martinab2 |
I looked after a few of these amps in the early 90s when I was working for an MF dealership. Don't remember any instability problems with this particular design - we had one pretty much permanently powered up in one of the listening rooms which was used with a wide range of different equipment.
They all ran really hot though, so 15 years on I would be looking for dried up filter caps and probably more cooked resistors. Iffy soldered joints on the pcbs are likely at this age as well.
It would be worth going through the whole amp and replacing the components that are past their sell by date - caps, power resistors etc. Check the tube sockets for the EL34s and the soldered joints between these and the pcb below.
Good luck |
|
|
| johndiy |
check resistors,electrolitics and joints :D
john |
|
|
| efflux |
OK, thanks for the replies.
I tested as many of the caps as possible but my multimeter can't go high enough uf to test them all.
The latest is that when I run each channel and test two valves in it. The amp runs with no motorboating. If I hook up 4 valves in any channel then I'm back to motorboating. Maybe this can give a clue for you to help me track down what is wrong. |
|
|
| efflux |
Another update here.
As I've said, when I run 4 valves I get the motorboating. There are two separate PCB boards with the valves attached and related circuitry such as cathode resistors and caps, grid resistors etc. These are connected to metal sections which also house the output transformers so when open these parts of the amp can be physically moved away from the main PCB. This movement cures the motorboating but with 4 valves in a channel I get a lower level hum (doesn't occur with 2). The amp has some kind of insulating spacers which are not there while I work and a lot of connections had shrink wrapped insulation which is obviously removed as well. It is possible some noise would be eliminated when rebuilding the amp but the motorboating does not disappear unless the valve PCBs with transformers are moved well away from the main PCB.
Apart from the electrolytics which I can not test with a simple uf on my multimeter, all the rest of the amps caps and resistors appear to be OK and I can not find any dodgy connections. A possible problem is burnt PCB around where the power resistors were but it is not serious. The PCBs seem a bit weak. some of the copper tracks around the power resistors were damaged and I had to bend the new resistor wires back across the PCB to make good contact but I can't see anything that now looks like a dodgy connection. |
|
|
| efflux |
EC8010.
You mention about the HT caps. I have discovered that I think they were replaced with a different value. The only thing I have to go on is that I have an invoice for the previous repair. The invoice says that 6 470uf 220v caps were replaced but in the amp these caps are actually 560uf 250v. I know sometimes you have to go higher because of lack of availability of the correct value. I make the assumption that 470uf caps were what was originally in there but they used higher value even although the invoice says different. Could the problems have stemed from this? I also have no idea about the make of these caps. All the other Electrolytics are Jamicon. Should I simply upgrade all the electrolytics? What are good quality or at least worth the cost? The 6 big ones may well be poor quality and I don't think Jamicon is that good either from what I remember.
Also. Oscilloscopes. I have considered getting one of these before. Any ideas about what is a good oscilloscope to buy? Not too expensive but a good make. Something that I might find on ebay for example.
This is my first valve hifi amp and I've had a taste of what a great sound it is. I'm prepared to go to a fair degree of expense and time just to get the sound back and better if possible which I'm sure it is. This amp had a gloriously powerful and detailed sound. |
|
|
| TJ |
You can try this number , not sure they still in the business or not ... they own the Michaelson & Austin .
Papworth Audio Technology,
32 Fendon Road, Cambridge, CB1 7RT
Telephone (01223) 416078 |
|
|
| efflux |
Hi TJ,
I telephoned Musical Fidelity who serviced the amp a couple of years ago. This is Michaelson's company. They wouldn't give me info. Basically they wanted me to send the amp for servicing. Judging by an invoice I have here from the service two years ago I will definitely not be sending it. Huge expense to replace a few caps, volume control and selector switch. I want to learn anyway.
One thing I can't understand at the moment is some resistors on the screen grids. They have colour code for 10 ohm yet all read around .9 ohm. I'm sure they are not the original resistors either but I have no schematic so I have no idea what was here originally. Two of these have gone. One was gone first time I looked in the amp (it had literally cracked apart). and another went up with the dodgy valve. They are tiny little wirewounds. At least that's what they look like to me judging by the one that has blown apart. I've not seen resistors like this before.
When I've said valve I mean tube of course. I presume non UK people know that we call them valves? It might sound confusing.
I've ordered Valve Amplifiers by Morgan Jones (I guess everyone will know we call them valves then). I need to learn more before I start doing anything else with the Odysseus. My guess at the moment is Filter Cap problems but I need to make sure all other components are in order.
I was able to completely rebuild my Fender Twin amp and do some repairs - change the balance control to bias etc, but I found that much easier to understand. Easier to see the wiring rather than looking at a circuit board and this Michaelson is a nightmare to work on unless you simply dismantle the whole thing and take it all out the chassis. |
|
|
| TJ |
| both side not working or just one side ? self bias or auto bias ? when u turn on the amp ,did the valve turn red ? or flashing ? |
|
|
| efflux |
Initially one side of the amp went. The fuse on that channel went. There are fuses on each channel as well as a main fuse. A valve had obviously gone as I found out when I turned the amp on again and a screen grid resistor blew on the dodgy valve and that channel fused again. A new valve I put in did not cause any fusing. I'm gradually working through what does what but there is more circuitry in here than on my Twin. The Odysseus auto biases itself but what is the difference between auto biasing and self biasing? My Fender Twin is fixed biasing meaning I adjust the bias myself.
There are no signs of glowing red valve plates or anything like that. The valves appear to be OK now. There was never any glowing or flashes. New and old valves are all bought from Watford valves. Exactly the same Harma STR E34L type. The problem now is motorboating. It seems something got damaged or something that was an accident waiting to happen has been pushed over the edge by whatever changes have occurred. I have made a few changes myself but not changed any values of anything or type except resistor wattage. I even put back the original resistors but no changes. The problem is that from simply looking at a few photos of Odysseus I know some resistors have been changed. Maybe not changed in type or even value but definitely different wattage handling. Also, the cathode capacitors were all changed. These are 100uf 100v electrolytic. These were changed in the repair job done by Musical Fidelity. However, the resistors that Musical Fidelity changed are not the ones that look completely different from photos. I should probably take some photos of what looks different from the photo I linked to in my first message.
Once I get this Valve Amplifier book I'll get better understanding so be able to be more specific. I will also have to dismantle the amp more to be able to see routing properly on the main PCB. I may also get a better capacitor tester so I can test the big caps.
Screen grid seems a dodgy area at the moment. I can section out an area where every adjoining component has had a problem including a valve but this may not necessarily be related to just one valve socket or valve. The new valve has not had a problem yet. The screen grid on my Fender Twin is a different arrangement. I need to look into exactly how the screen grid works. I know what it does in the valve but not exactly how the current flows for this. All I can do is give clues until I know more. I also need to work out routing around the filter caps and adjoining power resistors. I can't see this without looking at the back of the PCB again.
By the way, just to add here. I was told it is very dodgy to have no speakers connected to this amp. My Fender Twin is the same. It must never be turned on without speakers so I have been very careful that the Odysseus has never been turned on with no speakers connected. The speaker connections have been totally solid. |
|
|
| efflux |
Thanks Ben Goh but unfortunately that amp bears no similarity to mine.
My amp is ultra linear. Is this common place on hifi amps? My guitar amp is ultra linear but that is very rare with guitar amps. On the Odysseus the ultra linear taps from the output transformer are supplying DC to the screen grids from what my understanding is but there are other connections to the screen grid which are not present on my UL guitar amp. |
|
|
| TJ |
This is wht i guess , hope this help
this is a a common problem on the EL34 valve . when this tube near end of it life mostly will cos internal short circuit . destroy the cathode resistor and some power supply components . it depend how the designer to fussing the amp to protect the main and audio transformers . you cannot repair the amp without drawing the circuit diagram . please check
1, PSU heater for EL34 , how they connect ,components
2, check the cathode , components
3,Main supply
4, coupling cap |
|
|
| efflux |
OK, Thanks TJ.
I plan on making a circuit diagram. I am still waiting for the Valve Amplifiers book which should help me understand more.
Cathode resistors are OK but there are some other components there that I will get back here about once I have some diagrams. Heaters seem simple. They just come from the power transformer to the EL34s and then to the preamp tubes.
It certainly looks like the tube shorted at the screen grid. I am also sure that the fuses on each channel were not there on the early Odysseus amps. I can't see them in any pictures. However my amp has had these since built. I can tell they have definitely not been added. |
|
|
| TJ |
keep in mind , if valve failure/short circuit , the current will looking a path to ground !! something will have to burn !! you have to follow the route where the current found it way to ground !! remember in a valve amp half the main current is from the heater circuit and the other half is from the main . hopefully your output transformer still ok .
btw you wont get much help from the book you order . this amp too complicated and not reliable . My advice is send it to MF , get the repair done and play it for couple year then sell it . |
|
|
| planet10 |
| quote: | Originally posted by TJ
this amp too complicated and not reliable . My advice is send it to MF , get the repair done and play it for couple year then sell it |
Or gut it and start over (assumming the iron is OK.
dave |
|
|
| efflux |
OK, thanks a lot for the replies.
The amp is complicated all right but this kind of makes it intriguing.
It looks to me like the problem is that the amp is powerful with a hell of a lot of heat but lacks reliability due to underrated wattage on a lot of resistors. I don't even need to get my multimeter to see that there are problem resistors. Resistors that have clearly taken some severe heat. Clearly way under wattage from what I see in pictures. It's hard to tell if they have been in there from square one or not. They are not documented as having been replaced in the repair. The cathode resistors look like power surge resistors. Possibly 2 watts. They're coating is all cracking off. Someone had painted them with something as some kind of makeshift repair which seems odd to me. Why not just replace them with ones that are up to the job? My Fender Twin had been totally abused by previous users for over 30 years, played into the ground, blown output transformer, wrong speakers etc etc but I hardly needed to replace any resistors, only those near the valves. The cathode resistors on the Odysseus were reading OK but I did try replacing them. In pictures I see oblong ceramics that must be at least 7 watts. Same story with a couple of the resistors at the filter caps. The screen grid resistors look like around 1/2 watt. This doesn't seem right to me at all. Again in the pictures these are much bigger.
I don't trust Musical Fidelity. They were the last ones to handle this amp as far as I know and it doesn't look good. Even the invoice does not match what they did. One phone call to them was enough. The guy I spoke to said he had serviced these amps for five years at one stage but was not forthcoming when I asked him about specific details. He gave me the expected blurb about dangerous voltages etc but I know about this. If an owner electrocutes themselves it's not Musical Fidelities responsibility but they sure want to repair it all right to earn a few bucks. This is why I got into learning about these things. I wouldn't let anybody touch my Fender Twin now. Cowboys had already been at that.
I'm certainly no expert on valve amps but that's why I've ordered the Valve Amp book. It may not relate to this amp but it will help me build up a better picture on how amps work. It should maker it easier for me to sort out a schematic.
How would I know if the output transformers were damaged? Would they simply not work at all? All I can say is that everything runs OK when I have two valves in each channel and not four. No motorboating or anything. Are there any tests I can run on them, bearing in mind I have no specs for them other than what I can work out from the rest of the amp?
As long as the transformers are OK then the amp is good in my opinion. I won't mess further with it yet though. I'm going to do some research first then I may be back. |
|
|
| efflux |
These are some of the dodgy resistors. The top is a picture of one at the filter caps. Doesn't need much further explaining to see problems here. I was able to pull this out easily. The PCB burning is not quite as serious as it seems to look in this photo. The board was almost touching the resistor. The bottom is some of the Cathode resistors. Looks bad again but surely these resistors are not correct. Photos of other Odysseus amps show standard type oblong high watt ceramics. Bear in mind that the amp did actually work at first. This burning damage seems to be long term heat exposure. I don't think it just happened suddenly or at least not while I had the amp.
 |
|
|
| Steve M |
Efflux: My advice is give up on this amp!!
I owned an Odysseus for about 3yrs and although it sounded absolutely wonderful (powerful, valvey, musical and sweet to the max), I had nothing but trouble with it. It went backwards and forward to my technician ( a highly competent valve amp builder) about ten times over that period of time.
The main problem with the amp is that basically 'heat-and-pcbs don't mix'. It is a lot of amplifier housed in a small box, meaning HOT HOT HOT! The two vertically placed pcbs housing the EL34's virtually disintegrate and fall apart over time. Michaelson should not have placed the EL34 sockets directly on the circuit board, the heat transfer situation is not good!
If you are keen enough though, you could view it as a long term hobby project and completely rebuild and hardwire the whole amp - the beautiful sound of the amp may certainly be worth it!
Regards,
Steve M. |
|
|
| HollowState |
Hello efflux,
I'm new around here and this is my first reply. Although I'm not new to electronics. I'm 63 with over 40 years experience as a designer, builder & repair tech. I learned electronics on tubes when there were no transistors.
Motorboating: To combat motorboating it helps to understand that it is a form of low frequency oscillation caused by feedback. The source of which is not always apparent, but usually originates in the power supply. Specifically from a supply that has too high of an output impedance. This condition will not supply proper current to the circuit when loaded past a certain point. Typical causes are deteriorated filter capacitors, changed resistor values and even bad or improper DC rectification.
Another source of feedback is a lack of isolation between stages. This can happen through the power supply when a signal is seeing a high resistance to ground. (as in a bad filter cap) or is more closely coupled through a changed/lowered divider resistance. I've even seen a shorted filter choke cause motorboating.
So you will need to check the filter capacitors either by direct substitution or by looking at the ripple with an oscilloscope. Temporally bridging them with a new one of the proper rating will also help find a defective unit. The larger value caps installed are not the problem and are OK so long as they were not smaller. BTW, Jamicon's are not so good to begin with. They're cheap Asian imports.
Screen resistors: EL34/6CA7s have fairly low screen dissipation levels that are componded when running ultra linear. This is because they can actually run at higher then plate voltage do to being connected from taps on the output transformer that are closer to the B+ source then the plates themselves. A screen limiting resistor of 1000 ohms is typical and highly recommended. This goes double for most of the Chinese imports of today which IMHI are much inferior to the NOS American stuff.
An overheating screen will kill an otherwise good tube in short order. Especially when the tube is mounted horizontally as yours are which hurts proper convection cooling. A design layout that I do not approve of. You can visually check to see if your screen are running hot. In a darkened room, look very closely at the tube while it's running and peer inside the structure. If you see the screen glowing orange, it's running too hot. This may be difficult to see at first so keep looking. Increase the value ot the resistor in series with it. Doing so will not hurt the sound or reduce the power, but your tube will be much happier and live a lot longer.
Cathode resistors: Looking at the picture you posted, I see what appear to be either metal or carbon film types. If so, these are a poor choice since they will not stand any sort of overload. Not even for a few seconds unless they were being used to act as a fuse. If one resistor should open. the other tube in parallel with it must take up the load. Carbon composition resistors are very forgiving and will withstand an overload well, just like a tube.
I also tend to agree with Steve about the overall mechanical design of the amp. It's too cramped with too much heat from those horizontal fire bottles.
I hope this will help you.
Victor |
|
|
| martinab2 |
One of the difficulties in trying to repair an amp like yours is working out where to start. Current problems are likely to be the result of faults, general wear and ageing as well as poor previous repair work.
If there are lots of poor soldered joints around then further fault finding will be near impossible as the circuit will change every time you move something. Spend a couple of hours going over the entire amp and remaking soldered joints which either look suspicious or are likely to have been stressed (sockets, power resistors etc).
Replace anything which is obviously damaged or in poor condition e.g. screen and cathode resistors. Unless you are 100% sure of an electrolytic capacitor then replace it. Check wiring and grounding arrangements.
Start checking the voltages for power supply rails, plate, screen, grid and cathode of output tubes and drivers. Make sure that these are within safe limits for the tube and compare working and faulty channels. It can be helpful to sketch out a rough schematic and write in the voltages at each point. It that doesn't identify the problem then at least it will give other diyaudio forum members something to work with. |
|
|
| efflux |
OK, thanks very much for the opinions.
Thanks for the long message HollowState. Much appreciated. I have one question to ask if you don't mind. What about wirewound resistors on the cathodes? The original Odysseus definitely had large oblong shaped ceramic wirewound resistors on here, presumably chosen due to easy and cheap availability of high wattage. I can see these in some photos I have of an Odysseus dismantled. Possibly Michaelson Audio decided this was not a good idea and changed it on later models or somebody else changed it to the resistors you see in the pictures. Maybe whoever did it wanted to make sure the amp would be coming back for repair. Who knows. Also, if you look at the photos of the cracked up cathode resistors, they do have windings inside. I thought this meant they must be wirewound but I've rarely seen inside resistors anyway. However, I have seen resistors just like this that are sold as wirewound surge resistors rated at 2 watts I think. Am I right in presuming that, with all the heat issues in this amp, where a resistor is best higher wattage, then wirewound would be more robust due to availability of much higher watt.
After looking carefully at various photos (not many) of the Odysseus, it seems my version is a revised one, specifically the main PCB (ISSUE 02). The PCB is different. In mine you can see certain PCB tracks that are left overs from the first version.
If anybody else out there knows of anyone who has one of these amps, that could be very helpful but I'm guessing it's quite rare.
I got the Valve Amplfier book which is excellent. I guess this is going to be my tube amp bible. |
|
|
| vinylengine |
| Hi, I'm not sure what version of the amp I have (this is a piccy of mine on the kitchen table) and I know nothing about valve amps, but if you want some close up images of the inside I should be able to oblige? As long as it doesn't involve stripping down any further than the top covers and heat shields as mine is working fine and I don't want to upset the boat :whazzat: |
|
|
| efflux |
Hi vinylengine.
I'm glad to hear that someone else has one of these and without problems.
There is a picture of my Odysseus at that same site. Posted by the previous owner. The one with the paint stripped from the knobs to reveal the brass. The brass was all dirty. I've polished them to a high finish and will lacquer them. That was before the problems started. More serious issues to deal with now.
You could be of HUGE help to me because I am not entirely sure about a few things in the amp - what they were originally. Has your amp ever had any repairs done? Is it all original?
Taking off the parts that you mention is all that is needed. However the cathode resistors are not easy to see. Here is a picture to show you some of the parts I would like to know about. Is this your amp because it's at that same site as the photo you linked to and it looks like it's sitting on the same wooden surface?
The red circles are the thing I need to know most. They are resistors mounted vertically on the valve PCB. If you can tell me anything about these, any colour codes or markings it would be much appreciated. Photos would be cool. The green circles are 6 big electrolytic capacitors. Be careful if you open the amp. Probably you know the dangers anyway but these caps can hold big voltages even when the amp is off. You won't need to touch anything anyway. I think I know what value these caps are but any markings with uf or v would be a help. The blue circles are more electrolytic capacitors. There are 4 the same. Again, uf and v markings. There are another two small electrolytic capacitors that look just the same but very small. It's not absolutely necessary to know all these others but I want to make sure. Some of mine have been replaced.
The next thing would be helpful but these are not so easy to see. Here is a photo I found of an Odysseus valve PCB. It is upside down so the components I have marked are on the bottom when viewing an open Odysseus amp. The yellow circle is a resistor. Any info about these would be helpful. The purple circle is another electrolytic capacitor. You should be able read markings on this from the top of the amp without problems. There are 8 or these resistors and 8 capacitors. All the same. One on each valve.
The final parts are resistors connected at these squares marked with yellow circles from a photo of my amp, taken before fixing these parts. Any info about these would be further added help but I'm sure I have this right now anyway. In this picture your amp will most likely have different resistors here. Except for this photo I have used photos of other amps because they will probably look exactly like yours.
Also at the back of the main PCB you will see ODYSSEUS D.I.P ISSUE and a number. I don't think this is really that important. Just minor design changes.
If you can give me some of this info it would be very helpful. Thanks for offering to help. |
|
|
| vinylengine |
| Hi, Yeah that's my amp! I don't know if it's ever had any repairs done though? I bought it S/H and only went inside to clear out some of the dust before firing it up as it had been stood some time. I ended up wiping off some of the markings off the valves with a damp cloth (doh!) but they all used to say Michaelson Audio on them so even they may be original. I'll try to get some close-ups of the inside by the weekend - it might even help me if anything ever does go wrong; it always useful if you can see what component values are before they get burnt off (!) |
|
|
| efflux |
OK. Thanks.
The valves will probably all be original then. Also, from looking at the photo of your amp I can tell it's all original. Your amp will have the parts mine should have which probably explains why yours is still running.
The resistors encircled with red are the ones I really want to know about before anything else because on my amp these don't make sense and my problems point to a possible issue caused by these. The ohm value was reading so off from from the colour code as to not make sense and the colour code was not even difficult to read. The size of these doesn't make sense either, the smallest resistors in the amp when it looks like they should be of reasonable wattage handling. |
|
|
| HollowState |
Hi efflux,
Regarding cathode resistors and the use of wirewound units. This usually depends on what they are being asked to do by the designer. Please note, I'm unfamilier with the circuit of this amp as to whether it uses fixed or cathode bias. So here's what I mean.
If the output tubes are using cathode bias, the resistor needed will be on the order of several hundred ohms to produce the required negative voltage at the grid. (an educated guess assuming 50ma.) This will be a power level of a few watts and will require a wirewound part. And the original parts you described certainly sound like wirewounds.
If the tubes use fixed bias, and the resistors are there only to monitor bias current with a metering circuit, then they will be a much lower value (with lower voltage drop) and need only dissipate fractional wattage. So a non-wirewound type would suffice here.
Since the originals seemed wirewound, I would replace with the same type. They will not easily burn out if a tube shorts or runs hot from loss of controlling bias. And this reminds me of one of my pet audio peeves.
Some will say that wirewound resistors are bad for audio because they are inductive and should not be used. Ignore most of that. This is audio, not RF. If we were dealing with RF it would be different. But at audio frequencies any inductive reactance in a resistor of this small physical size in moot.
I looked again at the picture of the resistors you removed and they sure look like film types. Although I could, perhaps, be wrong. Most of my experience is with American made equipment. The exception being with some fine Bruel & Kjaer and Marconi stuff. European parts could be, or look different.
Film resistors (carbon & metal) are made on a ceramic core with a thin film deposited on the surface. A laser is used to trim the value by making a spiral cut along the length. To the unfamiliar, this may appear as a wound wire when it really isn't. If you have actually peeled off some wire then, of course, it is.
Victor |
|
|
| efflux |
OK. Thanks HollowState.
From what you say, the cathode resistors on my amp are not wirewound then.
Michaelson Audio definitely used wirewound resistors on the Original version of the Odysseus. You can see this clearly on one of the pictures but these are totally different from the cathode resistors on mine. They may have changed the design later.
My amp is cathode biased which ties in with what you are saying about the resistors.
You can actually get non inductive wirewounds but they are very expensive.
So I will replace the cathode resistors with larger wirewounds of the same value as already in there. Exactly the same resistors as in pictures of the original amp which will probably be 5 or 7 watts or thereabouts.
The most important thing is getting the screen grid resistors right. Vinylengine might be able to shed light on that because my resistors here don't make a lot of sense. Tiny wirewounds (these are definitely wirewound - I can see actual wire on a blown one) with ohm value at around a tenth of what the colour code says.
Any thoughts about where metal oxide is good to use? I did use these to replace a few resistors on my Fender Twin.
If you think the Odysseus is bad what about my Fender Twin guitar amp. Upside down tubes where all the heat flows up into the chassis! |
|
|
| HollowState |
When installing new screen resistors it isn't totally necessary to stay exactly and precisely with the value that may have been originally used. (Reread my earlier post on them.) They are there to limit screen dissipation. Typically 1000 ohms are used. This is not an arbitrary value on my part, but comes from published application data for EL34s in ultra-linear service.
If your tubes were 6L6s or 6550s or another beam pentode type this would, perhaps, be less important. But with EL34s it is. And especially with the Asian imports that are produced in different factories having varying degrees of tolerance and quality control. To use a low value, say around 50 or 100 ohms, would be of little value in reducing screen dissipation. They might serve as a parasitic suppressor or to couple to the plate in triode mode, but that is not the issue here.
Metal oxide resistors. Personally I'm not a big fan of them. Their application is better relegated to RF circuits and low power audio unless you want them to act like a fuse and not fire burn. Metal films will not take an overload and they open very quickly if stressed or pulsed. They're good for high frequency because of their deposited surface resistance. (skin effect)
Carbon composition resistors are just the opposite. You can stress them for short periods and they'll still work. Even if they smoke a bit, but not crack, they'll work. Can't do that with film resistors. I think the reason you see them used so often nowadays is because that's what is available. Comps are not made anymore and haven't been for quite a while. That's why NOS A-B, Ohmite and etc. command their high prices. Fortunately for us DIYers, there's enough still around, but not for manufactures.
Be carefull how many metal oxides you use in one amplifier. Many years ago, I replaced all the carbon resistors in one of my Ampex recorders with metal films thinking I would get better sound. I also replaced the paper caps with polyesters. What I did get was a gawd awful exaggerated high frequency response. That was a real ear/brain opener for sure. Never again.
I'm very familier with the older guitar amps from Fender and Ampeg. Many of them had their tubes upside down. But they were mounted on steel chassis that both sank and deflected the heat away from the wiring inside. Not circuit boards. As I remember, they were reliable with minimal heat related problems.
Victor |
|
|
| efflux |
Yes, my guitar amp has been reliable given the abuse it has had for 30 years. It has nearly all carbon composite resistors.
The Odysseus screen grid resistors have a colour rating of 10 ohm but are all reading around .8 or .9 ohm. I can't understand this. The colour coding is very clear. Maybe some more light can be shed on this from vinylengine. On many schematics I have seen the screen grid resistors are much higher. My Twin has 470 ohm here as far as I remember. |
|
|
| vinylengine |
OK, here goes. You'll have to excuse the photos; the amp is pretty dusty again and I had to rush the shoot. If you need better pictures or if I've missed anything just say and I'll take another look after the weekend. If you know the component number on the board it would help. As you can see though, there are some rather obvious differences...
Board issue 02
Valve PCB
Main Board
C32-C36
R61 + R62
 |
|
|
| efflux |
Thanks vinylengine. This is fantastic. You have supplied me with everything I need to know. :D
One question still remains though. On picture 7 down we appear to have a capacitor. It's connected to pin 4 on the valve. The screen grid. On my amp these look exactly like resistors with resistor colour code. The code that didn't make sense. Can anyone else shed some light on this? Are there small caps that look exactly like resistors with resistor type colour codes? |
|
|
| HollowState |
Good pictures!
However, #7 is a picture of an inductor, not a capacitor or resistor. See the notation L2 on the board. It says it is a 10 microHenry choke. The "µ" is the symbol mu, meaning micro. (one millionth of a Henry) This explains the discrepancy you were experiencing. Small inductors can often appear like resistors with color coded bands. You could replace them with same, or try a higher value resistor as suggested earlier. Your choice.
Victor |
|
|
| vinylengine |
I'm a little rusty but I seem to remember L is the symbol for an inductor so I guess the 10µ 'capacitor' is actually a 10µH choke?
*edit*
Oops it looks like Victor beat me to it :rolleyes: |
|
|
| HollowState |
The designer of this amp has obviously chosen to use small inductors to act as parasitic suppressors in much the same way that a small value resistor would. A perfectly acceptable enginering choice assuming the screens do not run hot. And this would depend on several factors such as the tubes themselves, the voltage used and even the location of the screen tapes in relation to the plates.
If you wish to keep the same parts there but have trouble finding replacements, you can hand make close approximations. Wind a single layer of enameled magnet wire tightly around a resistor and solder the ends to the leads. Use a high value resistor (5K or more) so you don't lower the Q to much. Chose appropriate size wire, perhaps #24 or 26, (just guessing here) and finish off with some heat shrink or plastic tubing.
Keep in mind that the actual value of the inductor in not critical. Any value moderately close to the original will serve you well.
Victor |
|
|
| efflux |
OK Thanks this is all extremely useful.
So to keep this amp the same or improve on what has been put in there (which I now realize are chokes) I am looking for 10uh 0.92A chokes (inductors) but what about their resistance values? What about the R8 on that choke? Is this resistance? Bear in mind that one these chokes had blown apart and was reading big resistance. Something seems not quite right. Why are they so tiny when the ones in the above picture are large? No doubt the motorboating is caused by the fact that I put low value resisters in there as replacements. |
|
|
| efflux |
Latest update here. Presumably this is what I'm looking for:
http://rswww.com/cgi-bin/bv/rswww/s...heID=uknetscape
This next inductor looks exactly like what is in the amp and I guess it is not too different. However it is discontinued anyway:
http://rswww.com/cgi-bin/bv/rswww/s...heID=uknetscape
Notice the difference in price. This appears right across this amp. Even on components that must be original. It looks like at one stage Michaelson Audio did a lot of cost cutting and in some cases this may have been to detriment. Maybe when the were heading for bankruptcy. The previous owner said the amp was only serviced once which means lots of cost cuts must have been made on the original amp.
Just out of interest. What is your serial number vinylengine? Mine is 040310.
By the way, I don't know exactly what an inductor does. I'll read up about that. |
|
|
| vinylengine |
Hi again,
My serial number is 0401XX so a little earlier than yours. I have no idea how many they made though? Also I'm not sure they went bankrupt? The Michaelson Audio that made the Odysseus ran alongside Musical Fidelity in the early 1990's, I think as a little side project for Anthony (and allegedly Tim de Paravicini) but I don't believe they went bust, they just evaporated...possibly due to the MF brand being more profitable? Information seems to be a little hard to come by... |
|
|
| efflux |
Yeah, maybe I'm wrong about that. It probably just wasn't profitable.
I apologize for posting those non working links by the way because on the RS site you have to go to the home page first.
I'm just in the process of fitting some new inductors. Just a couple of temporary ones of same value to replace the ones that broke. I'm almost certain the amp will be working after that. Then I'll replace them all with ones exactly like on the original Odysseus. The motorboating was almost certainly my doing due to the confusion over the inductors.
You're lucky with your amp vinylengine. Many of the parts are better than mine. I was wondering about some of the caps. Should I ditch the Jamicons? Since I have the amp open, now is the time to upgrade anything. It's not easy to get at everything inside this amp. As far as I can gather the filter caps are Panasonic but the rest of the electrolytics are Jamicons. Any thoughts about that? What are good caps? I don't mind spending a few quid to improve things including reliability. It doesn't cost much unless the filter caps need replacing. When I see bags of dirt cheap caps at Maplins with Jamicon written on them it makes me wonder. |
|
|
| martinab2 |
I'm not sure that Michaelson Audio was ever anything other than a branding/badge engineering experiment for Musical Fidelity. The Odysseus was made in the same factory as the rest of their products and appeared to share the same component suppliers and construction methods.
At the time, MF were full of creative energy and prepared to have a go at almost anything. Reliability, safety and longevity appeared to be well down the list of priorities though which made sure that their dealers service departments were always kept busy. Best of the bunch was the B1, their entry level class AB integrated transistor amp which had next to nothing in the way of protection. Once the customer had shorted its speaker terminals and blown the output transistors it would just smoke its mains transformer until the fuse blew! |
|
|
| vinylengine |
| I'm not suprised to hear they where made in the same factory; they even used the same blue paint for the decals... The build quality is pretty good on mine though with heavy castings and steel plate for the bodywork, solid brass controls, Alps volume and input selector and Rubycon/Elna capacitors splashed around. Nothing amazing but equal to some more expensive amps I've looked inside. Sound wise its difficult to knock the circuit design; it's not a million miles behind my £2000 Heybrook Signature pre/power on power and its got the same level of detail and transparency. It's also got just a touch of warmth to its presentation to betray the valves inside. It may be the best integrated amp I've heard? Mine seems to have lasted well too (crossed fingers) with no signs of repair despite being class A and running at a quoted 55 degrees Celcius (the casing gets very, very hot). It's definitely no B1 :shhh: |
|
|
| efflux |
| It's possible many of the amps are like mine with poorer parts. I did find one forum where people were talking about problems with an Odysseus. I can't see you having any problems vinylengine. My initial problems were purely due to underated wattage resistors. The motorboating will be my fault. I haven't turned the amp on yet but I've got temporary parts in which should solve the issue. I'll be turning it on in a few hours so fingers crossed. I think the heat issues are not a major problem as long as the components are up to it. I think the power valves are actually pretty well seperated from the rest of the amp behind metal plates. I've not had a valve amp until this one and the sound was tremendous and possibly not even as good as it could be given that I have some poorer components which are easily upgradable. I did see photos of a Musical Fidelity transistor amp somewhere on this forum with a power resistor that looked like it had gone on fire the damage was so bad. The same scenario as on my amp but a lot worse. That amp was fixed but something was clearly not up to scratch in the first place. |
|
|
| efflux |
The amp is now working. It's safe. No dodgy burning resistors or bad joints. Thanks for the info here. Otherwise, it might have taken me ages to realize about the inductors.
Since replacing those dodgy power resistors, fixing the bad joints and blown inductors, the amp has changed slightly. Before, it was loader but distorted slightly when turned really load (you couldn't turn it right up) but now the volume seems to make sense. No distortion at really load volume and the volume control doesn't turn the volume up so much. It adjusts more as how you'd expect.
I'm wondering about these Jamicon electrolytics. Would I notice any difference if I replaced them? There are 8 Jamicons at the cathodes. I do music production here so I actually have 24 bit 96 Khz audio coming from a top end computer soundcard. This sounds immense through the Odysseus but maybe minor differences between caps may have an effect. Any opinions here would be appreciated. The amp is in bit's so it's easy for me to do any upgrades now rather than later. |
|
|
| HollowState |
Congratulations on getting the amp up and running.
| quote: | | I'm wondering about these Jamicon electrolytics. Would I notice any difference if I replaced them? |
This really depends on how sensitive you are to small minor changes. I know people who claim to hear differences in things like fuses, power cord plugs and even different labels on the same brand of tubes. (production runs?) This is way beyond my sensitivities, and perhaps a blessing in disguise. This is something you will just have to try for yourself.
While I feel that Jamicon is a cheap brand of capacitor, they do work. And I've always believed that if something is working, don't bother fixing it. However if you'd like to replace them with a better make for prevention's sake, I guess now is the time. There's certainly plenty of electrolytic brands to choose from. Nichicon, Nippon, Black Gate, Sprague, Siemens, Mepco/Electra, Ero, Rifa, Mallory and many more that I can't think of.
Victor |
|
|
| efflux |
| OK. I guess this is something I will just have to try. I might replace some of the Jamicons and see if there is any difference. It depends a lot on the rest of your gear. I'm testing the amp with an iPod at the moment but when I hook it up to my computer soundcard which is a pro production card then I hear detail that is way beyond the iPod or most stand alone CD players. Then it goes much further when I listen to my own recordings because I'm hearing them at 24 bit 96 Khz which is another leap above CD quality so it's possible components in the amp could make a slight difference that I can hear but my speakers are not top notch. Linn Keosas. However they seem to be pretty good with the valve amp. The screen grid inductors and cathode resistors are all to be replaced. The dodgy looking cathode resistors in the photos I posted are back in because they do actually work. I'll replace these parts first and take it from there. |
|
|
| efflux |
Hi again.
I haven't changed any parts on the amp yet but some parts will be replaced. Dodgy resistors and screen grid inductors all to be replaced to something more like the original and possibly replace the electrolytics. Everything else seems OK. However I have a couple of questions before I start going through this amp in detail. Some of my repairs to get it working are only temporary fixes. There are a couple of other things to fix like a broken phono socket. I will need to pull the main PCB out again. I want this amp so that when I close it I won't be opening it again for years, if possible.
The power transformer makes a low humming sound. Is this normal with a transformer of this size? This does not get into the audio output except when at full volume, then you do hear a faint hum. However this is at volume that you would never turn to. I'm just slightly concerned about this hum and what may be causing it.
The other point is about electrolytic values. For most of the electrolytics I can get values as original. Some of the replaced electrolytics are slightly higher value. As far as I am aware this isn't a problem. However I have noticed the 6 big power electrolytics have been replaced with 560uf 250v when the originals in vinylengine's amp are 470uf 200v. I think mine were originally 470uf 220v judging from an invoice I have. I'm just wondering what values I should use. I can get 470uf 250v (or 200v) Vishay BC, Epcos, Cornell Dubilier or Panasonic electrolytics (Panasonic is actually what I have but 560uf - these caps may be OK but I won't know until I start stripping the amp down). These are the cap makes I can source in the UK at these higher values. What difference can it make upping uf or volts in these caps as has been done? What is more important - keeping uf or volts near original? Are there any problems associated with taking these values higher? For the other smaller electrolytic caps I can use just about any make, even Black Gates although very expensive.
Any advice appreciated. Thanks for helping. |
|
|
|
