Last year there was some interest in developing a retrofit board for Hafler/Adcom boards:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=64226&highlight=
http://www.diyaudio.com/forums/showthread.php?s=&threadid=64226&highlight=
Overwhelmed new guy questions
My venerable DH-220 seems to be developing more mechanical transformer hum. Google got me to diyAudio, and forums search got me to this thread.
I've made one pass through the thread, but I think it will take another to digest everything! The first read was enough to have me asking questions, if somebody can manage to humor yet-another new guy:
1. I don't actually use the DH-220's power switch. It is plugged into a switched outlet on the DH-110 (WCF == Wife Complexity Factor precludes anything else). Am I in danger of cooking this switch from the turn-on current draw of the DH-220?
2. If I'm going into the DH-220 to add rubber washers for the transformer mounts, it sounds like I should be checking values on the power supply caps and replacing if they have lost value (or upgrading to more if I also want to put a higher-rated bridge and soft-start in), and adding smaller-value caps in parallel regardless of keep/replace/upgrade. Have I got all that correct?
3. I see C405 on the schematic (effectively across the secondary), but would a snubber cap in parallel with each diode in the bridge be helpful (I've seen it in other power supply schematics)?
Thanks for answers and newbie tolerance!
My venerable DH-220 seems to be developing more mechanical transformer hum. Google got me to diyAudio, and forums search got me to this thread.
I've made one pass through the thread, but I think it will take another to digest everything! The first read was enough to have me asking questions, if somebody can manage to humor yet-another new guy:
1. I don't actually use the DH-220's power switch. It is plugged into a switched outlet on the DH-110 (WCF == Wife Complexity Factor precludes anything else). Am I in danger of cooking this switch from the turn-on current draw of the DH-220?
2. If I'm going into the DH-220 to add rubber washers for the transformer mounts, it sounds like I should be checking values on the power supply caps and replacing if they have lost value (or upgrading to more if I also want to put a higher-rated bridge and soft-start in), and adding smaller-value caps in parallel regardless of keep/replace/upgrade. Have I got all that correct?
3. I see C405 on the schematic (effectively across the secondary), but would a snubber cap in parallel with each diode in the bridge be helpful (I've seen it in other power supply schematics)?
Thanks for answers and newbie tolerance!
OK... I have a Question that goes back a little ways... there was a mention of a schottkty diode in series with the bridge rectifier + and - to eliminate noise in the p/s . could you not parallel all 4 sides of the bridge ??? and what amp and volt specs would you use???
Thanks for your patience for such a late response .
Elwood.
Thanks for your patience for such a late response .
Elwood.
AM Roberts and eYoung
If you study the schematic for the DH220 you will see that each electrolytic capacitor has its own small film bypass capacitor. There are no "quick fixes" or simple mods worth doing to the DH-220. They already have been incorporated into the circuit.
The jury is still out on whether faster diodes or snubber capacitors on the diode bridge are worth the effort in improved sound. If someone can point me to a credible well designed and implemented A-B test of this mod I would like to read it.
The only mod I have found worth the effort (and expense) for the DH-220 is to replace the original 10K uF filter caps with something like the 26K uF caps from Musical Concepts. This extra capacitance with 2 uF film bypass caps (and 6800 ohm bleeder resistors for safety) can make an observed and heard difference in the sound. This is worth doing. But, the extra capacitance puts a much greater load on the ON/OFF switch and diode bridge on start up.
This extra PS capacitance is right on the edge of requiring a huskier power switch -- if you can find one. Many who add extra PS capacitance also use a soft start circuit to gradually (over 1 second or so) to turn on their amplifiers. I use such a device built up in the metal case from an old AT computer power supply (the large version) that uses a 3-4 second time delay relay with 117 VAC field coil and a 50 ohm power resistor in series with the incoming AC. Circuits of this type are all over the place here, do a couple of searches.
Some people use a CL-40 Thermistor - Type CL, soldered in series with the incoming AC supply and placed in free air to aid cooling. This thermistor slows down (microseconds) the initial AC surge enough to provide adequate protection to the power switch. Some argue that a thermistor reacts too slowly to the varying power demands of an A/B type amp circuit, some say it does not make a difference in the sound. Later versions of the DH-500, for example, used a thermistor. I have used one in my DH-220 with extra capacitance and heard no difference and have modified two other DH-220s for friends who also heard no difference with the thermistor in place.
The DH-220 came along after all the many mods for the DH-200 were published. The DH-200 needed the mods, the DH-220 does not.
Another mod mentioned in threads here is to place a 470 uF electrolytic in parallel to the existing 10,000 uF caps in the PS. These smaller newer electrolytics can make a difference in the sound of the old tired caps in most old Haflers, and I have heard this difference. Here is a picture of such a mod:
If you study the schematic for the DH220 you will see that each electrolytic capacitor has its own small film bypass capacitor. There are no "quick fixes" or simple mods worth doing to the DH-220. They already have been incorporated into the circuit.
The jury is still out on whether faster diodes or snubber capacitors on the diode bridge are worth the effort in improved sound. If someone can point me to a credible well designed and implemented A-B test of this mod I would like to read it.
The only mod I have found worth the effort (and expense) for the DH-220 is to replace the original 10K uF filter caps with something like the 26K uF caps from Musical Concepts. This extra capacitance with 2 uF film bypass caps (and 6800 ohm bleeder resistors for safety) can make an observed and heard difference in the sound. This is worth doing. But, the extra capacitance puts a much greater load on the ON/OFF switch and diode bridge on start up.
This extra PS capacitance is right on the edge of requiring a huskier power switch -- if you can find one. Many who add extra PS capacitance also use a soft start circuit to gradually (over 1 second or so) to turn on their amplifiers. I use such a device built up in the metal case from an old AT computer power supply (the large version) that uses a 3-4 second time delay relay with 117 VAC field coil and a 50 ohm power resistor in series with the incoming AC. Circuits of this type are all over the place here, do a couple of searches.
Some people use a CL-40 Thermistor - Type CL, soldered in series with the incoming AC supply and placed in free air to aid cooling. This thermistor slows down (microseconds) the initial AC surge enough to provide adequate protection to the power switch. Some argue that a thermistor reacts too slowly to the varying power demands of an A/B type amp circuit, some say it does not make a difference in the sound. Later versions of the DH-500, for example, used a thermistor. I have used one in my DH-220 with extra capacitance and heard no difference and have modified two other DH-220s for friends who also heard no difference with the thermistor in place.
The DH-220 came along after all the many mods for the DH-200 were published. The DH-200 needed the mods, the DH-220 does not.
Another mod mentioned in threads here is to place a 470 uF electrolytic in parallel to the existing 10,000 uF caps in the PS. These smaller newer electrolytics can make a difference in the sound of the old tired caps in most old Haflers, and I have heard this difference. Here is a picture of such a mod:
Attachments
Snubbers or fast diodes
Here's two articles on this topic I've found helpful:
Rick Miller, "Measured RFI Differences Between Rectifier Diodes...," The Audio Amateur, 1/94, p. 26..
and,
http://www.hagtech.com/pdf/snubber.pdf
Miller advocates the use of diodes with soft recovery characteristics. The devices used also happen to be high speed, which is what many DIY'ers have latched onto. But, it's the soft recovery that leads to reduced ringing.
Hagerman suggests that rather than buying special diodes, just dampen the oscillation by using a proper snubber.
Of the two, I like Hagerman's approach the best.
Dick West said:
Here's two articles on this topic I've found helpful:
Rick Miller, "Measured RFI Differences Between Rectifier Diodes...," The Audio Amateur, 1/94, p. 26..
and,
http://www.hagtech.com/pdf/snubber.pdf
Miller advocates the use of diodes with soft recovery characteristics. The devices used also happen to be high speed, which is what many DIY'ers have latched onto. But, it's the soft recovery that leads to reduced ringing.
Hagerman suggests that rather than buying special diodes, just dampen the oscillation by using a proper snubber.
Of the two, I like Hagerman's approach the best.
"Last year there was some interest in developing a retrofit board for Hafler/Adcom boards:"
Yes, but what if I want more power than a PLH puts out (25W~30W)?
I want something that will replace all the Hafler boards (except the 120 series).
I was thinking of something along the lines of the Musical Concept board, but with a bootstrap current source for the Vas (like an AKSA) and an RC filter with zener feeding a resistor for the LTP (I prefer the resistor to most active CCS).
Yes, but what if I want more power than a PLH puts out (25W~30W)?
I want something that will replace all the Hafler boards (except the 120 series).
I was thinking of something along the lines of the Musical Concept board, but with a bootstrap current source for the Vas (like an AKSA) and an RC filter with zener feeding a resistor for the LTP (I prefer the resistor to most active CCS).
Weinstro,
My comment:
was meant to relate to the sound of the DH-200/220 power amps. The Rick Miller article you referenced was a sidebar to the POOGE article about DAC Modifications. The Hagerman article begins with reference to the need of snubbers when "an oscillatory circuit must be damped." Do these relate to a Hafler power amp?
Over the years I have read about the efficacy of fast diodes and snubbers with digital equipment; but, my memory tells me that many more technically astute that I am have said these enhancements to the typical analog PS in amplifiers are not useful as a 60 hertz PS is "too slow" to make them worthwhile.
My memory (sometimes faulty) tells me that even Nelson Pass said special diodes are not required in an amplifier and modifiers such as John Hillig was very slow to add them to his mix of Hafler mods available. Did John add them because they really make a difference or because they are an item to improve sales? Was Mr. Pass inferring that they are not needed in one of his Class A amps but might be worthwhile in a Class AB amp with varying power demands?
I know that bench tests with an o'scope and signal generator can show before/after results of some mods but are they hearable from a typical Hafler amp in a typical home music system? Many modifiers state they do, but their comments may be based on "builder's bias" and not on carefully controlled A-B tests.
I want to learn and even have an IXYS bridge and a pair of identical Hafler amps. One of these days I may install the IXYS bridge in one and compare their sounds, but I would be more motivated to do so if there were a credible A-B test to refer to.
Thanks for your comments about this.
My comment:
was meant to relate to the sound of the DH-200/220 power amps. The Rick Miller article you referenced was a sidebar to the POOGE article about DAC Modifications. The Hagerman article begins with reference to the need of snubbers when "an oscillatory circuit must be damped." Do these relate to a Hafler power amp?
Over the years I have read about the efficacy of fast diodes and snubbers with digital equipment; but, my memory tells me that many more technically astute that I am have said these enhancements to the typical analog PS in amplifiers are not useful as a 60 hertz PS is "too slow" to make them worthwhile.
My memory (sometimes faulty) tells me that even Nelson Pass said special diodes are not required in an amplifier and modifiers such as John Hillig was very slow to add them to his mix of Hafler mods available. Did John add them because they really make a difference or because they are an item to improve sales? Was Mr. Pass inferring that they are not needed in one of his Class A amps but might be worthwhile in a Class AB amp with varying power demands?
I know that bench tests with an o'scope and signal generator can show before/after results of some mods but are they hearable from a typical Hafler amp in a typical home music system? Many modifiers state they do, but their comments may be based on "builder's bias" and not on carefully controlled A-B tests.
I want to learn and even have an IXYS bridge and a pair of identical Hafler amps. One of these days I may install the IXYS bridge in one and compare their sounds, but I would be more motivated to do so if there were a credible A-B test to refer to.
Thanks for your comments about this.
DJK,
The PLH retrofit thread was doomed because the Hafler power trans was too high voltage and, as you point out, the power available was too low.
There must be thousands of broken Haflers gathering dust that a suitable universal retrofit board could rejuvenate. Maybe even a board with a chip front end to make things simpler. A lot of the broken Haflers are languishing neglected because their owners are not technically astute enough to repair them, so a retrofit board kit simple to build could be the best strategy for them to use.
I am not technically astute enough to design one but would be an avid kit builder if boards and BOM were available.
The PLH retrofit thread was doomed because the Hafler power trans was too high voltage and, as you point out, the power available was too low.
There must be thousands of broken Haflers gathering dust that a suitable universal retrofit board could rejuvenate. Maybe even a board with a chip front end to make things simpler. A lot of the broken Haflers are languishing neglected because their owners are not technically astute enough to repair them, so a retrofit board kit simple to build could be the best strategy for them to use.
I am not technically astute enough to design one but would be an avid kit builder if boards and BOM were available.
Dick West said:
Any inductive element (e.g., a transformer) feeding a capacitive element (e.g., a capacitor filter bank) will form an oscillatory circuit. The bridge rectifier creates sharp on/off excitation. The Hafler has this, as does every device with a linear power supply.
Miller's article was an eye-opener because it showed very clearly that commonly accepted circuits can have unknown characteristics. A lowly 1N4004 is plenty fast for 60 Hz circuits, but it's the sharp cutoff that causes the ringing.
Hagerman basically says, yes, it rings. So, damp the circuit. He then gives an example where the ringing is basically eliminated with a 20 cent part.
Dick West said:
For a while, these ultrafast recovery diodes were quite the rage. If I were a vendor of parts, I would rather sell 4 diodes at $4 each than a ceramic cap at 50 cents.
Dick West said:
Good point -- can I hear electrical noise at 100 kHz? Probably not, and I doubt they generate subharmonics in the audio range. But, they do contribute to losses in the circuitry. If I had pushed up the capacitance in the filter bank, and was close to the ratings of the diode bridge, I would go ahead and put a snubber on it.
Dick West said:
My suggestion would be to follow the measurement technique in Miller's article, and follow the approach specified in Hagerman's article.
Hafler Mains Switches
I am astonished that despite previous suggestions about these switches that they are still raising questions. It would take just a short time to solve this once and for all.
this can be done in two simple ways .. Throw out the old switch and make a metal plate to cover the hole using self tapping screws and then affix a readily available
double 15 amp toggle type switch to the plate.. secondly from the same electronic distributor use a double pole type with light inside, which will simple mean a little work with a file to open up the oblong hole but will be more in keeping with the original setup. Its that simple.
Its not a good idea to run the power amp off the preamp outlet because that is not the purpose of those outlets they are for lower current equipment..also I see from Dick Wests modified Hafler the original twin flex mains cable is still in use,meaning if you want more beef put a decent size cable in the amp as already suggested in other
writings on this forum You seem determined to put the cart before the horse..meaning in the States you have nominally 110volts supplies that means
twice the current of our mains in the UK which means twice the cable losses.
About diodes the 1N4004 is actually quite noisey compared to the larger 25AMP types this is probably because the larger diodes and bridges have a larger forward capacitance giving a flatter noise distribution but this is relative to the current being switched, most people place much smaller caps across the main reservoirs of 2to 4.7
mfds and somtimes 0.1mfds across that to snub out unwanted grunge from the bridge ..some engineers feel that this is the best all round solution because it is hard sometimes to evaluate differences, schottkys and the like seem to be preferred in equipment that generates a variety of different noise like CD players
etc, where improvements do seem to work
regards
I am astonished that despite previous suggestions about these switches that they are still raising questions. It would take just a short time to solve this once and for all.
this can be done in two simple ways .. Throw out the old switch and make a metal plate to cover the hole using self tapping screws and then affix a readily available
double 15 amp toggle type switch to the plate.. secondly from the same electronic distributor use a double pole type with light inside, which will simple mean a little work with a file to open up the oblong hole but will be more in keeping with the original setup. Its that simple.
Its not a good idea to run the power amp off the preamp outlet because that is not the purpose of those outlets they are for lower current equipment..also I see from Dick Wests modified Hafler the original twin flex mains cable is still in use,meaning if you want more beef put a decent size cable in the amp as already suggested in other
writings on this forum You seem determined to put the cart before the horse..meaning in the States you have nominally 110volts supplies that means
twice the current of our mains in the UK which means twice the cable losses.
About diodes the 1N4004 is actually quite noisey compared to the larger 25AMP types this is probably because the larger diodes and bridges have a larger forward capacitance giving a flatter noise distribution but this is relative to the current being switched, most people place much smaller caps across the main reservoirs of 2to 4.7
mfds and somtimes 0.1mfds across that to snub out unwanted grunge from the bridge ..some engineers feel that this is the best all round solution because it is hard sometimes to evaluate differences, schottkys and the like seem to be preferred in equipment that generates a variety of different noise like CD players
etc, where improvements do seem to work
regards
Re: Hafler Mains Switches
In the US you can't use a double switch in this application (115VAC line). This was pointed out to me after I had written an article for AX and I checked it out with a friend at the ARRL -- in the US you would be switching the ground -- in the ROW you are switching two hot lines.
In one of my 2kW linear ham band amplifiers the relay switches two lines -- but these are on the 220VAC line.
humble said:
In the US you can't use a double switch in this application (115VAC line). This was pointed out to me after I had written an article for AX and I checked it out with a friend at the ARRL -- in the US you would be switching the ground -- in the ROW you are switching two hot lines.
In one of my 2kW linear ham band amplifiers the relay switches two lines -- but these are on the 220VAC line.