Good idea....I'll play with that. I've also thought of adding that common mode conduction prevention transistor the M-1.0t has to the M-500t.
The M-500t has less heatsink, 2 rails, and no common mode transistor.
The M-1.0t has more heatsink, 3 rails and a common mode transistor.
So, the M-1.0t runs cooler; I've been wondering if adding a common mode transistor to the M-500t would help her run cooler, just haven't had the time to try it yet.
The M-500t bias network has that thermistor, so when things heat up the bias drops back down.
The M-1.0t doesn't have this, so it can climb higher than you want it to.
Later series, like the TFM had a much better bias network, a sort of sliding bias that was rock steady at any temp.
The M-500t has less heatsink, 2 rails, and no common mode transistor.
The M-1.0t has more heatsink, 3 rails and a common mode transistor.
So, the M-1.0t runs cooler; I've been wondering if adding a common mode transistor to the M-500t would help her run cooler, just haven't had the time to try it yet.
The M-500t bias network has that thermistor, so when things heat up the bias drops back down.
The M-1.0t doesn't have this, so it can climb higher than you want it to.
Later series, like the TFM had a much better bias network, a sort of sliding bias that was rock steady at any temp.
Hi Rich,
Good idea. You will need to back off the bias transistor or you may end up with not enough bias control range. After you get that figured out, use a lower value bias control and a series resistor. This is just to make things easier to adjust. That would be a good idea for the supply rail adjust control too.
I think that "rule" (do not change the circuit!!) is for good reason. For all my time servicing, an ailing unit that was modified normally returns to it's normal robust service after you return the device to factory trim. Small component changes don't bother me, the problem is that most people do not know when to stop, or how to select components that will function properly. Some times the PCB is badly damaged also. Between lifted traces, burned off copper, deep knife cuts / cracked boards and holes now drilled out too large, some items are just too badly damaged to be worth repairing.
A properly executed circuit modification by technicians who do not have access to the warranty network is rare. Most are butcher jobs that might even endanger the equipment owner. When I see a good job done, properly engineered and executed, I am very pleasantly surprised. I always tell the owner that their technician did very good work in those cases. Credit where credit is due.
-Chris
You mean that transistor that often shorts in an output failure that most techs can't find?
Good idea. You will need to back off the bias transistor or you may end up with not enough bias control range. After you get that figured out, use a lower value bias control and a series resistor. This is just to make things easier to adjust. That would be a good idea for the supply rail adjust control too.Given that the M 1.0 rarely hits the third tier, you may find the extra heat is merely due to the higher first rail voltage, then when it hits tier two, you really get into more dissipation. Still, the M-500 does not get that hot by comparison with other amplifiers. Once you match those transistors, you may be able to reduce your bias current for the same or lower THD figures.
I have never actually sat down to compare the different amps for bias behavior. That is due to the fact that, as a warranty shop, we return the product to it's normal behavior for that model. To invest the time and effort to do something that the distributor or manufacturer does not want done is not possible. Not if we wanted to retain our warranty authorized status. Any changes we ever worked on was in response to an actual fault or weakness in design. This only while working with an engineer while taking great care not create any other problems. This mind set is pretty ingrained in me after so many years doing warranty service.
I think that "rule" (do not change the circuit!!) is for good reason. For all my time servicing, an ailing unit that was modified normally returns to it's normal robust service after you return the device to factory trim. Small component changes don't bother me, the problem is that most people do not know when to stop, or how to select components that will function properly. Some times the PCB is badly damaged also. Between lifted traces, burned off copper, deep knife cuts / cracked boards and holes now drilled out too large, some items are just too badly damaged to be worth repairing.
A properly executed circuit modification by technicians who do not have access to the warranty network is rare. Most are butcher jobs that might even endanger the equipment owner. When I see a good job done, properly engineered and executed, I am very pleasantly surprised. I always tell the owner that their technician did very good work in those cases. Credit where credit is due.
Well, it don't think it's sliding bias. It does have better control though. However, the heat sinks are larger and this can also lower the tempco of your bias current. I have a TFM-753x I haven't used yet. I might be interested in having a closer look at this guy to figure some stuff out. The only problem is that I do not have any service info on it any more. That, I'll need before attempting anything. This amp is brand new in box, it's intended for a surround sound room. Should be fun.
-Chris
anatech said:Hi Rich,
You mean that transistor that often shorts in an output failure that most techs can't find?
Yep, that's the transistor. I'm going to add it to the M-500t I'm working on now and see if preventing common mode conduction saves much heat. The stock alignment pots are pretty crappy, open frame. I replace then with enclosed that are much smoother, but you're right, they DO have way too much range. A static res and a pot of half value would be smoother.
I completely agree with this. Several of the amps I bought second hand had been 'juiced' from a 74VDC rail to 85VDC. For one thing, the top tier caps are 80WV, so it's flirting with disaster at the onset. Also, the V gain stage devices are 150V and are now being asked to swing 170V. If these units ever did provide more power, they most likely clipped or had greatly increased distortion.
I always up the caps WV to exceed what the new rails will be, and replace the V gain stage with devices to handle the new voltage swing maximums. Being pre-drivers and drivers, they won't sag the rail like the outputs will under heavy current sourcing, so I like them to meet the maximum expected voltage swing.
IN the M-1.0t, for example, the stock rails are 30, 60 and 100VDC. The stock gain stage is comprised of 200V units (well within the 150VDC swing).
The mk II rails are 33, 68 and 115VDC. The new gain stage uses 230V devices.
Similarly, for the M-500t (74VDC rail) I increase the gain stage devices from 120V and 150V to 160V and 180V
I think the class G rail switching is what makes this whole plan possible. If it were a static rail, a 15VDC rail jump would be , I believe, ill-advised and possibly destructive.
Being a rail switcher, the low rail's 3VDC bump, and the most often in use (mid) rail's jump of 8VDC are, IMO, relatively benign.
The whole idea was NOT to make these amps louder. That IS a side effect, but the main thrust was to source more current for better cone motion control.
I really liked the sonics of the stock M-500t, but thought it was a bit weak in the bass dept. My contention being that ability to source current has more to do with 'tight' bass than 'damping factor' (unless damping factor is a horrible 2, there's decimal point differences between infinite damping and a factor of 4; it becomes a numbers game that can be manipulated with feedback.)
That was the genesis of the mk II project, better sound, not louder.
Yes, it's not a TRACKING bias like the new tube amp he's making....that's why I said sliding (which isn't technically correct either)......Well, it don't think it's sliding bias. It does have better control though. However, the heat sinks are larger and this can also lower the tempco of your bias current. I have a TFM-753x I haven't used yet. I might be interested in having a closer look at this guy to figure some stuff out. The only problem is that I do not have any service info on it any more. That, I'll need before attempting anything. This amp is brand new in box, it's intended for a surround sound room. Should be fun.
-Chris
The gist being that in the network that drifts more with heat, the heatsink mounted bias transistor is directly responsible for the V span at the top and bottom of the bias network. As it heats up, the whole network span climbs.
In the more stable version, that heatsink transistor tickles the base of a second (off heatsink) transistor, which spans the network emittor to collector. It's the second transistor that's responsible for the bias network span, and there's less drift.
Hi Rich,
-Chris
That is a very common misconception. Those open trim controls can be very good quality, and they do have robust contacts. Sealed controls still do oxidize, but at a little slower rate. What you really have to watch for is a more lightly built control where the contact surface is smaller and possibly under less pressure. Multi-turn controls are not required here, not are they recommended. Their contacts are smaller and under less pressure. I'm sure you can buy really good ones, but no one does!
Absolutely! This is the better way to accomplish the goal.
Those amps either drifted up on their own, or there are some amazingly stupid technicians out there. Sorry, I won't mince any words here, running supply caps over there ratings is just plain stupid. Doesn't buy you any extra power either (in dB).
I still worry about the dissipation in all the parts. Not worthwhile if you convert your extra power into a difference in dB.
That is really a function of output resistance, or damping factor. The "t" designation has resistance inserted in series with the speaker output. Want to make the bass tighter? Short out this resistor. That's an easier modification and requires no changes to the amplifier or power supply circuits.
But, the fact is that control over the woofer is a function of the output impedance of that amplifier. That is why most amplifiers that use negative feedback has a falling damping factor characteristic as you increase the frequency. The amount of available feedback is falling, and if there is an output inductor, it's impedance is rising with frequency. In other words, the resistance of the source (the amp) is rising as the frequency increases. The ratio of a standard eight ohm load, divided by the output impedance of the amplifier gives you a figure of merit called "damping factor". Therefore, bass control is tied up into the damping factor figure no matter how you look at it. Damping factor matters up to some value where the resistance of the connections and wire limit how low you can make the output impedance. In other words, within reason.
-Chris
Yes, that series resistor was suppossed to give 'rolling, tube-like bass' IIRC. I don't like it personally. The M-4.0t and 1.5t have it, but the M-500t and M-1.0t do not, and those are the models I chose to work on (primarily because they're easier to take apart....no vertical boards, just an open layout).
Yes, I agree damping factor matters, I'm saying that, with the human nervous system's response of about 12-14 milliseconds (on a good young day) the difference in decay time, or overhang (or whatever you want to call it) in a driver seems to matter little if the amp's damping factor is 20 or 200
http://www.audioholics.com/education/amplifier-technology/damping-factor-effects-on-system-response
so to an extent, the damping factor spec has become over-emphasized, I believe.
Hi Rich,
A loudspeaker is a current operated device. That is what we are both getting at. The ability of an amplifier to deliver to deliver that current as required determines how that speaker will move. Internal resistance, no matter how is created, will reduce the ability of an amplifier to excite a speaker and to damp it's movement. It does this by overcoming the back emf of that speaker. These effects would be in the first msec., and I believe what you are concerned with.
A higher supply voltage may assist in this, but it is not necessary to raise the supply voltages to accomplish the same thing. The effect of the feedback network along with the rest of the amplifier circuit determines how well that speaker is controlled, all other things being equal. For instance, you could take the feedback from across the internal crossover connections, like remote sensing in a power supply. This would reduce or eliminate the effects of wire and connection resistances because they would then be within the feedback network. Neat eh?
Don't forget to that some speakers are more difficult for an amplifier to control than others. Depends on the crossover and even how hot the voice coil normally runs.
I'm pretty sure that other modifications to the Carver amplifiers would solve that particular issue. What you don't like seems to be precisely what Bob Carver designed them to be.
-Chris
A loudspeaker is a current operated device. That is what we are both getting at. The ability of an amplifier to deliver to deliver that current as required determines how that speaker will move. Internal resistance, no matter how is created, will reduce the ability of an amplifier to excite a speaker and to damp it's movement. It does this by overcoming the back emf of that speaker. These effects would be in the first msec., and I believe what you are concerned with.
A higher supply voltage may assist in this, but it is not necessary to raise the supply voltages to accomplish the same thing. The effect of the feedback network along with the rest of the amplifier circuit determines how well that speaker is controlled, all other things being equal. For instance, you could take the feedback from across the internal crossover connections, like remote sensing in a power supply. This would reduce or eliminate the effects of wire and connection resistances because they would then be within the feedback network. Neat eh?
That's why I maintain that it is important within reason.
It really does between 20 and, say 80 or a little more. I would argue that a damping factor between 200 and 1,000 is far less important.
Don't forget to that some speakers are more difficult for an amplifier to control than others. Depends on the crossover and even how hot the voice coil normally runs.
I'm pretty sure that other modifications to the Carver amplifiers would solve that particular issue. What you don't like seems to be precisely what Bob Carver designed them to be.
-Chris
That would be a nice thing, to extend the feedback loop to the crossover internals. I'm running active crossover though.
It's not that I don't like it, but rather, in comparison, the M-500t seems (to my ears) to be a tish soft in the lower registers. Overall I really like the sound, and am very taken with the elegance of the concept. After the mk II mod, and for my money, I couldn't be happier with the dynamics.
I'm back from vacation, and sent an email out to all who expressed interest in having their unit's mk II upgraded.
If everybody (10) on my list goes ahead with it, I'll be booked up until the 1st week of April, so if you get an email, let me know as soon as you can if you:
1. Are staying on the list
2. Want to be bumped down on the list
3. Are no longer interested
Thanks, and Happy new Year!
If everybody (10) on my list goes ahead with it, I'll be booked up until the 1st week of April, so if you get an email, let me know as soon as you can if you:
1. Are staying on the list
2. Want to be bumped down on the list
3. Are no longer interested
Thanks, and Happy new Year!
Hi TNRabbit,
The M-500 is not the best match for those speakers, a Lightstar or Sunfire is on the other hand. Something along the lines of a pair of M 1.5L might work well, but my money would be on TFM-75 or PT-2400 (close to the same beast). There you get both sound quality and next to infinite current drive.
Another amplifier you could consider would be the Nakamichi PA7 (modified) or PA-7II. Threshold amplifiers would be another good match. Old Krell would sound too sterile, new ones might be okay.
Hi Rich,
The earlier M1.0, no "t" would be a good starting point for you. Examine the differences between that and an M 1.0t to get an idea how to "back step" your M-500t. I can't remember if an M-500 was ever produced. If so, that might actually be something closer to the sound you are looking for.
-Chris
Interesting combination.
The M-500 is not the best match for those speakers, a Lightstar or Sunfire is on the other hand. Something along the lines of a pair of M 1.5L might work well, but my money would be on TFM-75 or PT-2400 (close to the same beast). There you get both sound quality and next to infinite current drive.
Another amplifier you could consider would be the Nakamichi PA7 (modified) or PA-7II. Threshold amplifiers would be another good match. Old Krell would sound too sterile, new ones might be okay.
Hi Rich,
That means even better enhancements, you maintain all your damping factor to the driver. The normal passive crossover is an evil only intended to make things work, not to make things work well. I wish more speaker manufacturers would make electronic crossovers for their speakers, or at least publish the parameters required.
Then remove the "t" modifications. You don't like the "tube transfer function modified" direction. Hey, that sounds like it could be the later "TFM" series. Tube Transfer Modified!
The earlier M1.0, no "t" would be a good starting point for you. Examine the differences between that and an M 1.0t to get an idea how to "back step" your M-500t. I can't remember if an M-500 was ever produced. If so, that might actually be something closer to the sound you are looking for.-Chris
yes, I'm VERY happy with the sound now that I've eliminated the passive crossovers. Snap your head back dynamicsanatech said:
That was my inititial direction, remove the T mod from the M-500t (it's only a handful of parts and a couple of different R values). One thing led to another and I ended up with the mk II, which I'm extremely pleased with.
The downstairs bar area (AKA my workbench) has been having a new ceiling put in due to water damage. They finished today, so I'll be back to 'work' (i have upgrades booked until April). I'm going to take 'unit 1' (my guinea pig) out and mess around with a couple of ideas I have re the bias network. I'm also going to measure the gain stage transistor temps as you suggested (I have baseline temps from the unmodified version) and will see if running the gain stage off of the top rail puts them in harms way. The outputs seem to run 5 to 10 F hotter, and I'm hoping the gain stage shows comparable results or I'll have to think about snap-on heatsinks (probably a good idea anyway with the stock unit.
Hi Rich,
Isn't it amazing how much damage a passive crossover can do? Mind you, getting everything right in an active crossover is too difficult and time consuming normally. Unless you can get the crossover frequencies, slopes, level changes and any eq. from the manufacturer, I would avoid doing this. However, designing your own system can be made much easier by going active right off the start.
Good luck on your projects.
-Chris
Yup. That's about right. More acoustic power now as well, right?
Isn't it amazing how much damage a passive crossover can do? Mind you, getting everything right in an active crossover is too difficult and time consuming normally. Unless you can get the crossover frequencies, slopes, level changes and any eq. from the manufacturer, I would avoid doing this. However, designing your own system can be made much easier by going active right off the start.
You went in two different directions though. Removing the "t" "voicing" is one change done for sound quality only. Increasing the power is an entirely different change. One where you hope you don't mess up the sound quality. I'm not saying you did, it's just that higher voltage circuits can require different parts and sometimes circuit changes. There is where trouble awaits!
I'm sorry to hear that. Loo failure?
In some amps, they run too hot anyway.
That's something you will never hear me complain about. Drilling some holes below can help airflow a great deal.
Good luck on your projects.
-Chris
It's been awhile (I'm a bit slow ) but I finally got around to taking temp measurements at the gain stage.
From before, I knew that the heatsinks ran about 5-10 degrees F hotter. Some had asked me to check temps at the gain stage.
I chose to test the M-500t because:
1. It's what I was working with at the time
2. with 2 rails it's a bit less efficient than the M-1.0t's three rails
Measurements:
1. stock parts, stock rail:--(74VDC rail with 120V parts)
2. mk II parts, stock rail:--(74VDC rail with 160V parts)
3. mk II parts, mk II rail:--(95VDC rail with 160V parts)
-------------------Stock--------mk II/stock--------mk II/mk II
gain stage:-----112-------------100-----------------106
predrivers:------106-------------100-----------------106
drivers:----------120-------------104------------------114
It seems that the higher rail and new gain stage parts DON'T cause overheating in this stage either, which is good news!
From before, I knew that the heatsinks ran about 5-10 degrees F hotter. Some had asked me to check temps at the gain stage.
I chose to test the M-500t because:
1. It's what I was working with at the time
2. with 2 rails it's a bit less efficient than the M-1.0t's three rails
Measurements:
1. stock parts, stock rail:--(74VDC rail with 120V parts)
2. mk II parts, stock rail:--(74VDC rail with 160V parts)
3. mk II parts, mk II rail:--(95VDC rail with 160V parts)
-------------------Stock--------mk II/stock--------mk II/mk II
gain stage:-----112-------------100-----------------106
predrivers:------106-------------100-----------------106
drivers:----------120-------------104------------------114
It seems that the higher rail and new gain stage parts DON'T cause overheating in this stage either, which is good news!
Hi Rich,
Your temperature readings show that the new package types are what is saving the day. However, I've always felt that the parts ran too hot originally. So I guess it's no worse, but I was hoping for an improvement here. See if you can find some small snap on or glue on heat sinks for those. As you can see, the parts marked "driver" transistors are the most sensitive.
Can you test them under load after the amp has been running at power? Try a 4 ohm load.
-Chris
Your temperature readings show that the new package types are what is saving the day. However, I've always felt that the parts ran too hot originally. So I guess it's no worse, but I was hoping for an improvement here. See if you can find some small snap on or glue on heat sinks for those. As you can see, the parts marked "driver" transistors are the most sensitive.
Can you test them under load after the amp has been running at power? Try a 4 ohm load.
-Chris
Hi Rich,
Good show!
The original parts are hot enough to cause board damage after some time, and that was noticed years ago. Now everything is even older.
If you can get the part temperatures down, you will be making a meaningful change. It's great that you were already considering that.
-Chris
Good show!
The original parts are hot enough to cause board damage after some time, and that was noticed years ago. Now everything is even older.
If you can get the part temperatures down, you will be making a meaningful change. It's great that you were already considering that.
-Chris
-hello all...
-ive read alt about you mr anatech
-i can not quote as of just yeti...i would like to add the carver ss mono-blocks to the tfm75 and her pro series sister...perhaps the 600~2...i do not have that much belief in LS [opinion]
-any of those amps are capable of giving more thn the ALS spkr systems can handle
-hope i didnt interupt too bad
-ps; hi sir richp !!!
-ive read alt about you mr anatech
-i can not quote as of just yeti...i would like to add the carver ss mono-blocks to the tfm75 and her pro series sister...perhaps the 600~2...i do not have that much belief in LS [opinion]
-any of those amps are capable of giving more thn the ALS spkr systems can handle
-hope i didnt interupt too bad
-ps; hi sir richp !!!
Hi maddmaxxx,
Understand that using all the power output from an amplifier is not what you want to do. High audio power in a speaker is mostly converted to heat energy. The higher the voice coil temperatures get, the less efficient the actual speaker (or driver) gets. Every speaker will have a maximum SPL they can create in a given volume of air, modified by the interaction with the structure of that space. This is determined by Xmax and the piston area of the drivers. Adding more energy to this only serves to decrease that maximum SPL and heating the glue to a point where everything lets go. I have seen foil burned off a ribbon. Not pretty, but expensive.
Why are people fixated on how much power a speaker will "handle" and how much power you can get out of a single amplifier??
-Chris
Oh, oh .... I'm in trouble now!
Sorry, I am having trouble deciphering this. Can you restate your comment please?
The TFM-75 is a stereo amplifier. Quite a good one actually. It will drive ALS speakers and almost anything to beyond the breaking point. You don't even need to bridge it for this. I'll bet your AC supply gives out before this amplifier does. These are one of my favorite amplifiers that Carver Corp. ever released.
I think the PT-2400 was the Pro equivalent to the TFM-75. For home use, I would use the TFM-75 if the choice was mine.
I'm not sure what you are referring to here. Can you explain "LS" please? Do you mean Lightstar by chance? If so, I did own one of these and serviced some as well.
The TFM-75, PT-2400 and Lightstar are all capable of destroying the ALS speaker system. As I mentioned earlier, other factors will come into play that might limit your power output.
Understand that using all the power output from an amplifier is not what you want to do. High audio power in a speaker is mostly converted to heat energy. The higher the voice coil temperatures get, the less efficient the actual speaker (or driver) gets. Every speaker will have a maximum SPL they can create in a given volume of air, modified by the interaction with the structure of that space. This is determined by Xmax and the piston area of the drivers. Adding more energy to this only serves to decrease that maximum SPL and heating the glue to a point where everything lets go. I have seen foil burned off a ribbon. Not pretty, but expensive.
Why are people fixated on how much power a speaker will "handle" and how much power you can get out of a single amplifier??
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.