Howdy y'all,
I just got some old Melcor mic pre cards in the mail. They've got the 1731 opamp and Melcor input and output trannies, but to my suprise, they say "Made in UK." I thought Melcor was American, not that I mind, just a little confused. They are model AM-20, of which I can find zero info anywhere. I know about the AM-27, but has anybody heard of AM-20?
Thanks!
Anyone know about Melcor AM-20 Mic Pre's? Made in UK???
As a matter of fact, they are from Port on ebay! I hooked one up with jumper cables last night and indeed it did sound great, I can't wait to rack these up... So these are prototypes of the AM-20? I still can't find any info on the AM-20, was it pre-AM-27? Does that make this pre-pre-pre-API?
Here's the pic:
http://cgi.ebay.com/2-Vintage-Melcor-AM ... dZViewItem
Thanks.
Here's the pic:
http://cgi.ebay.com/2-Vintage-Melcor-AM ... dZViewItem
Thanks.
-
The Scum
- moves faders with mind
- Posts: 2746
- Joined: Thu Jul 03, 2003 11:26 pm
- Location: Denver, CO
- Contact:
Those look like Vector-type prototype cards...generic proroboard circuit cards made to go in a Vector rack chassis. You can solder up whatever curcuit you want on the raw cards. I think you can still get the cards from Vector today.
It's quite possible that the naked cards were made in the UK, but then bought by Melcor and turned into preamp modules in the USA.
Byron Jacquot
It's quite possible that the naked cards were made in the UK, but then bought by Melcor and turned into preamp modules in the USA.
Byron Jacquot
I beleive Byron is correct. The card stock itself was made in UK, but the components are Melcor. That is basically an API 312 preamp and should sound killer. The Melcor opamps are slightly smoother, and not quite as aggresive sounding as the api 2520. In many cases I like the 1731 better because of that. The beyer transformer may change the tone a little verses the Jensen tranformer API used, but probably not in a better or worse way. Just different.
I almost bought those at least 20 times over the last few months. Just yesterday when I saw them sold, I mentioned to a guy what a great deal someone got. So Congrats.
Joe
I almost bought those at least 20 times over the last few months. Just yesterday when I saw them sold, I mentioned to a guy what a great deal someone got. So Congrats.
Joe
Thanks! I emailed the seller and he said they were just UK perfboards and hardware. The schemo looks a little different from the API, I think this may have been designed for unbalanced out because the output tranny seems like an after-thought and the output actcually goes through an electrolytic coupling cap to the transformer, which makes zero sense to me, haven't decided if I should do a full on API conversion or keep them in their original historic (to me) condition.
I was looking at this auction several times a day for a while too, then when I realized I already had a usable power supply in an old Valley People rack, I had to go for it. Thanks for the congrats!
Anyone know where I can get an AM20 schematic??
I was looking at this auction several times a day for a while too, then when I realized I already had a usable power supply in an old Valley People rack, I had to go for it. Thanks for the congrats!
Anyone know where I can get an AM20 schematic??
Search for AM 27 it should be the same. If you can't find it, PM and I will send it to you. There is not to many ways to connect up the 1731, and two tranformers to be honest with you. The API and Melcor are probably a lot closer to the same than you think. Schematics may look different, but I would bet they are not really. Look at a 325 and 312. they look alot different on paper. But with not a lot of problems, you can see that the are identical circuits with a slight difference in components value...
Oh yeah, the cap and resistor before the transformer was to try and keep dc from the audio path. Same on API...
Oh yeah, the cap and resistor before the transformer was to try and keep dc from the audio path. Same on API...
Hmmm...
I'm looking at an API 325 sschematic right now:
http://www.diyfactory.com/projects/project2/API_312.JPG
...And the signal goes straight out of the opamp into the transformer, no cap in series connecting them. I thought output transformers were air-gapped so they could handle DC flowing through them while only passing AC. Though the API has a cap on the - side of the output xformer, maybe that's a way of accomplishing the same thing while avoiding running the ouput through the cap. (?)
Most of the differences, besides that one, are subtle, for instance a cap going in series with a resistor, instead of in parallel. But there is one big difference and that is on the AM20 there is a positive feedback thing going on: a line going from the 1731 output back into the + input through a resistor. All I have is Port's hand-drawn schematic so I don't know the value, maybe it's really high and keeps the circuit from oscillating (??), maybe it's a mistake and it's supposed o go to the - input. I've got some tracing out to do. Thanks for the help!
I'm looking at an API 325 sschematic right now:
http://www.diyfactory.com/projects/project2/API_312.JPG
...And the signal goes straight out of the opamp into the transformer, no cap in series connecting them. I thought output transformers were air-gapped so they could handle DC flowing through them while only passing AC. Though the API has a cap on the - side of the output xformer, maybe that's a way of accomplishing the same thing while avoiding running the ouput through the cap. (?)
Most of the differences, besides that one, are subtle, for instance a cap going in series with a resistor, instead of in parallel. But there is one big difference and that is on the AM20 there is a positive feedback thing going on: a line going from the 1731 output back into the + input through a resistor. All I have is Port's hand-drawn schematic so I don't know the value, maybe it's really high and keeps the circuit from oscillating (??), maybe it's a mistake and it's supposed o go to the - input. I've got some tracing out to do. Thanks for the help!
Well wow....where to begin. To begin with I am not all that knowledgeable about electronics I just pretend to be...
All the 2520/1731 and its variants that I know of work with a feedback circuit into the opamp. They may not always be drawn in a way that it is obvious though. This is how the opamp works. It tries to make the two inputs even in the same voltage. So by feeding back in the voltage on one side, the otherside kicks it up a notch to match. This is pretty much the fundamentals of the circuit. I am sure someone else can do a more thorough explination. If Port drew it as a + and not - I am guessing it is a mistake.... but that said, by the definition, it could probably work on either + or -, but that is out of my range of knowledge.....BRIAN?
API 2503 transformers are not gapped. They are basically 4 identical windings around a plastic bobbin. 1 winding gets used a primary and the other 3 as secondaries. It really does not matter in that design which is which.
The cap in the 312 schematic (47pf I think) is controlling the amount of high freqs that can pass. On a 325 card I think that cap is like 120pf. In a line amp situation this is done to not change the incomming signal. It should have all of its tone already added. So the bandwith is like -3db at 38Khz. Mic amps need to control some of that freqs or the tone becomes harsh. By changing that cap it drops the responce to about -3db at 28khz. Rolling off the harsh high end. That is not the same cap that is in the output section of the 325 to the transformer. The 312 goes straight into the transformer in the original design. I imagine you could do the same on your transformers. When converting my 325 cards into 312 cards, I remove the cap and resistor and go straight into the tranny. It seems to open the sound up some......
Wow...I guess I have been hanging around the electronic experts to long...
All the 2520/1731 and its variants that I know of work with a feedback circuit into the opamp. They may not always be drawn in a way that it is obvious though. This is how the opamp works. It tries to make the two inputs even in the same voltage. So by feeding back in the voltage on one side, the otherside kicks it up a notch to match. This is pretty much the fundamentals of the circuit. I am sure someone else can do a more thorough explination. If Port drew it as a + and not - I am guessing it is a mistake.... but that said, by the definition, it could probably work on either + or -, but that is out of my range of knowledge.....BRIAN?
API 2503 transformers are not gapped. They are basically 4 identical windings around a plastic bobbin. 1 winding gets used a primary and the other 3 as secondaries. It really does not matter in that design which is which.
The cap in the 312 schematic (47pf I think) is controlling the amount of high freqs that can pass. On a 325 card I think that cap is like 120pf. In a line amp situation this is done to not change the incomming signal. It should have all of its tone already added. So the bandwith is like -3db at 38Khz. Mic amps need to control some of that freqs or the tone becomes harsh. By changing that cap it drops the responce to about -3db at 28khz. Rolling off the harsh high end. That is not the same cap that is in the output section of the 325 to the transformer. The 312 goes straight into the transformer in the original design. I imagine you could do the same on your transformers. When converting my 325 cards into 312 cards, I remove the cap and resistor and go straight into the tranny. It seems to open the sound up some......
Wow...I guess I have been hanging around the electronic experts to long...
-
Family Hoof
- buyin' a studio
- Posts: 877
- Joined: Mon Apr 19, 2004 5:30 pm
- Location: NYC
- Contact:
Interesting find! I have a few comments...
AFAIK, Melcor got into the calculator business a long time ago, dropped their audio line, and never looked back. API (with ex-Melcor employees) kept making compatible stuff. From about 1970 until maybe a decade later I guess, API 312, 325 cards and 550 EQs were very popular building blocks in many consoles. For example, many of the consoles around NYC like Datamix and the stuff at Hit Factory and other big facilities were custom jobs which used a lot of API/Melcor items, but weren't API consoles. Given the popularity I believe there were at least 2 or 3 other companies making their own 312 and/or 325 style cards, and you had other 550 compatible EQs and such. I wasn't around then, so I can't guarantee the accuracy of this brief history, but it's the jist of what I've been able to gather from talking to people who were.
The Api 312 circuit is about as simple as an op amp-based mic pre can get. Are you looking at one now? I will try to explain what each component does...
The mic input transformer, usually an AP2622, provides a balanced input and has two primary coils which can be connected in either series or parallel, depending on the the desired input Z or step up. C6 and C7 are shunt capacitors connecting from each of the primaries to ground. Because they're very low value they only offer a low Z path (lower impedance than the transformer input) to high frequencies, above the audio band. These frequencies might exceed the bandwidth of the 312 circuit, and cause distortions. Beause of this the 1000pF caps are used to shunt them to ground, essentially a low-pass filter.
Next we see that the secondary of the input transformer unbalances the signal by connecting one side of the winding to ground. The other side goes straight into the +, or non-inverting, input of the op amp. The series combination of R4 and C5 are in parallel with the input transformer secondary as another type of filter, to damp any high frequency ringing and smooth out frequency response anomolies. This is often referred to as a zobel network.
Now we see that the output of the op amp is fed back into the -, or inverting, input, but through a bunch of components. R3, C4, R2, and C3, as well as an external pot, comprise the feedback network. This is how we manipulate the amount of gain the op amp gives. Each components function will be discussed in a minute. First some background.
For our purposes all we need to know about an op amp is this. It has two equa inputs, which get summed, amplified, and output as a single signal. One input inverts the incomming signal and one does not. The op amp also has an 'open loop' gain - that is the amount of gain it provides without any feedback loop (the loop is open instead of closed). This is usually somewhere from 85 - 120dBu. much more than we need. If we close the loop with a straight piece of wire, feeding the amplified output back to the inverting input it cancels out the amplification, and the output is unity gain. In reality, no op amp would stable enough to operate with a straight wire closed loop. Also, we want to be able to vary the gain form unity on upwards. So our feedback loop needs a network of passive components to control it.
First the resistors. R3 and R2 comprise a voltage divider. The full voltage goes through the series resistor, R3, then some of it sees a low Z path to ground through the shunt resistor, R2, and the rest goes to the non-inverting input as feedback. Because not all of the signal is fed back, we get some gain. The amount of gain depends on the ratio of R2 to R3. If we connect a potentiometer in series with R2, often about equal in value to R3, then we can change the voltage divider ratio and thus have full ocntrol over gain. the low value R2 prevents the shunt part from going all the way down to 0. If all of the feedback was shunted we'd essentially have open loop gain. Not only is this too much gain, as mentioned, but it's rare that an op amp will operate stabilly at open loop gain. there will undoubtedly be distortion. The relatively high value of R3 drops the current of the feedback voltage down to something that is sensible for an input - since the output of the op amp has a very high current to drive the load.
Now the caps. Series C4, whose value is usually at or below 100pF, compensates for phase shift of very high frequencies at the output by bypassing (providing alternate low Z path around) the series resistor. Its value is specific to the op amp and should be exact. Because if the inverting input signal is out of phase with the non-inverting it won't cancel. It will be regenerative feedback instead of degenerative and cause oscillations. HF oscillations will then cause unwanted distortions down in the audio band. Shunt C3 is a DC blocking cap. Its value must be very large so that low frequencies are not rolled off. Essentially a sort of high-pass filter. Because it needs to be an electrolytic cap to get that high in value, it will have a polarity, a DC bias (unless bi-polar/non-polar type of lytic). We measure the polarity of the DC offset at this poiint in the circuit and install C3 accordingly. You see, there is always inevitably some DC offset voltage at the output of the op amp. We don't want DC in our feedback loop, AC only (AC being the audio), because DC through a pot makes a scratchy sound, and an unequal amount of DC at the inputs would not cancel, it'd be amplified, compromising headroom and creating a larger DC offset at the output. Because transformers, unless they're designed with a gap in the core (most aren't), can be damaged or compromised in their performance from DC, most mic pres would have another large blocking capacitor between the op amp output and output transformer primary winding. API didn't think this would be a problem with the 312, so there isn't one.
Now to finsing the circuit... We see the AP2503 output transformer rebalances the signal. one side of the primary winding gets the unbalanced (aka single-ended) output signal and the other side is referenced to ground. The three secondary windings can be connected in a number of ways depending on what you're driving... but that's another story.
Finally, the two diodes, CR1 and CR2, and the two capacitors, C1 and C2 are there to provide local power supply decoupling for the op amp. Because the 312 is presumably being connected in parallel with a bunch of other amps to the same power supply, we want to somehow isolate the power it gets to avoid problems with noise and cross-talk. the diodes prevent reverse polarity of the DC power supply voltages. Since this is unlikely to happen, most circuits would have a low value current limiting resistor in their places. The caps serve two functions- 1) low Z path to ground, shunting high frequency noise (anyhting above 0Hz, aka DC, if their values are high enough), and 2) charge up to provide the op amp with its own little resevoir.
Anyway, I hope this little essay is helpful to someone. As you can imagine, explaining or substituting all the tech lingo I used would've doubled the length of this piece. You can goolge search all this shit anyway. To all the technically-inclined readers, sorry if my dumbing-it-down offended you.
AFAIK, Melcor got into the calculator business a long time ago, dropped their audio line, and never looked back. API (with ex-Melcor employees) kept making compatible stuff. From about 1970 until maybe a decade later I guess, API 312, 325 cards and 550 EQs were very popular building blocks in many consoles. For example, many of the consoles around NYC like Datamix and the stuff at Hit Factory and other big facilities were custom jobs which used a lot of API/Melcor items, but weren't API consoles. Given the popularity I believe there were at least 2 or 3 other companies making their own 312 and/or 325 style cards, and you had other 550 compatible EQs and such. I wasn't around then, so I can't guarantee the accuracy of this brief history, but it's the jist of what I've been able to gather from talking to people who were.
The Api 312 circuit is about as simple as an op amp-based mic pre can get. Are you looking at one now? I will try to explain what each component does...
The mic input transformer, usually an AP2622, provides a balanced input and has two primary coils which can be connected in either series or parallel, depending on the the desired input Z or step up. C6 and C7 are shunt capacitors connecting from each of the primaries to ground. Because they're very low value they only offer a low Z path (lower impedance than the transformer input) to high frequencies, above the audio band. These frequencies might exceed the bandwidth of the 312 circuit, and cause distortions. Beause of this the 1000pF caps are used to shunt them to ground, essentially a low-pass filter.
Next we see that the secondary of the input transformer unbalances the signal by connecting one side of the winding to ground. The other side goes straight into the +, or non-inverting, input of the op amp. The series combination of R4 and C5 are in parallel with the input transformer secondary as another type of filter, to damp any high frequency ringing and smooth out frequency response anomolies. This is often referred to as a zobel network.
Now we see that the output of the op amp is fed back into the -, or inverting, input, but through a bunch of components. R3, C4, R2, and C3, as well as an external pot, comprise the feedback network. This is how we manipulate the amount of gain the op amp gives. Each components function will be discussed in a minute. First some background.
For our purposes all we need to know about an op amp is this. It has two equa inputs, which get summed, amplified, and output as a single signal. One input inverts the incomming signal and one does not. The op amp also has an 'open loop' gain - that is the amount of gain it provides without any feedback loop (the loop is open instead of closed). This is usually somewhere from 85 - 120dBu. much more than we need. If we close the loop with a straight piece of wire, feeding the amplified output back to the inverting input it cancels out the amplification, and the output is unity gain. In reality, no op amp would stable enough to operate with a straight wire closed loop. Also, we want to be able to vary the gain form unity on upwards. So our feedback loop needs a network of passive components to control it.
First the resistors. R3 and R2 comprise a voltage divider. The full voltage goes through the series resistor, R3, then some of it sees a low Z path to ground through the shunt resistor, R2, and the rest goes to the non-inverting input as feedback. Because not all of the signal is fed back, we get some gain. The amount of gain depends on the ratio of R2 to R3. If we connect a potentiometer in series with R2, often about equal in value to R3, then we can change the voltage divider ratio and thus have full ocntrol over gain. the low value R2 prevents the shunt part from going all the way down to 0. If all of the feedback was shunted we'd essentially have open loop gain. Not only is this too much gain, as mentioned, but it's rare that an op amp will operate stabilly at open loop gain. there will undoubtedly be distortion. The relatively high value of R3 drops the current of the feedback voltage down to something that is sensible for an input - since the output of the op amp has a very high current to drive the load.
Now the caps. Series C4, whose value is usually at or below 100pF, compensates for phase shift of very high frequencies at the output by bypassing (providing alternate low Z path around) the series resistor. Its value is specific to the op amp and should be exact. Because if the inverting input signal is out of phase with the non-inverting it won't cancel. It will be regenerative feedback instead of degenerative and cause oscillations. HF oscillations will then cause unwanted distortions down in the audio band. Shunt C3 is a DC blocking cap. Its value must be very large so that low frequencies are not rolled off. Essentially a sort of high-pass filter. Because it needs to be an electrolytic cap to get that high in value, it will have a polarity, a DC bias (unless bi-polar/non-polar type of lytic). We measure the polarity of the DC offset at this poiint in the circuit and install C3 accordingly. You see, there is always inevitably some DC offset voltage at the output of the op amp. We don't want DC in our feedback loop, AC only (AC being the audio), because DC through a pot makes a scratchy sound, and an unequal amount of DC at the inputs would not cancel, it'd be amplified, compromising headroom and creating a larger DC offset at the output. Because transformers, unless they're designed with a gap in the core (most aren't), can be damaged or compromised in their performance from DC, most mic pres would have another large blocking capacitor between the op amp output and output transformer primary winding. API didn't think this would be a problem with the 312, so there isn't one.
Now to finsing the circuit... We see the AP2503 output transformer rebalances the signal. one side of the primary winding gets the unbalanced (aka single-ended) output signal and the other side is referenced to ground. The three secondary windings can be connected in a number of ways depending on what you're driving... but that's another story.
Finally, the two diodes, CR1 and CR2, and the two capacitors, C1 and C2 are there to provide local power supply decoupling for the op amp. Because the 312 is presumably being connected in parallel with a bunch of other amps to the same power supply, we want to somehow isolate the power it gets to avoid problems with noise and cross-talk. the diodes prevent reverse polarity of the DC power supply voltages. Since this is unlikely to happen, most circuits would have a low value current limiting resistor in their places. The caps serve two functions- 1) low Z path to ground, shunting high frequency noise (anyhting above 0Hz, aka DC, if their values are high enough), and 2) charge up to provide the op amp with its own little resevoir.
Anyway, I hope this little essay is helpful to someone. As you can imagine, explaining or substituting all the tech lingo I used would've doubled the length of this piece. You can goolge search all this shit anyway. To all the technically-inclined readers, sorry if my dumbing-it-down offended you.
Awesome! Great explination. I can only explain it at even a dummy level of knowledge. Ive been populating the 325 cards, and doing the conversion this last week, and am getting around them pretty good. But some of the techinical jargan around the input is a little confusing to calculate for me.
Anyway, glad you came around to help describe the circuit.
Anyway, glad you came around to help describe the circuit.
-
Family Hoof
- buyin' a studio
- Posts: 877
- Joined: Mon Apr 19, 2004 5:30 pm
- Location: NYC
- Contact:
I was describing the 312 circuit. I've never done a 325 conversion. What op amp are you using? What input transformer are you using? Looking at the schematic, I'd approach it like this...
- remove C3 and R2, and solder shorts in their places. Then jumper pin 7 to COM. - C5 can also be removed or replaced with a zobel network if your input transformer needs it.
- R1 gets replaced with whatever value you need to load the secondary of your input transformer with. Now connect the hi and lo of the secondary to pins % and 4, respectively.
- forget about pin 9
- connect the wiper and CCW end of your reverse log pot (10k or 22k (25k nowadays) depending on op amp) to pin 8. then connect a 200ohm resistor and 330uF/6.3V BP electrolytic cap in series with the pots other terminal and connect the other side of the cap to COM.
- R3 gets changed to either 10k or 20k depending on what op amp and pot value you used.
- C6 may need to be changed as well
- upgrade C4 to a 470uF/16V BP electrolytic cap... OR just remove it
- change C1 and C2 to the biggest value that'll fit at that voltage.
- remove C3 and R2, and solder shorts in their places. Then jumper pin 7 to COM. - C5 can also be removed or replaced with a zobel network if your input transformer needs it.
- R1 gets replaced with whatever value you need to load the secondary of your input transformer with. Now connect the hi and lo of the secondary to pins % and 4, respectively.
- forget about pin 9
- connect the wiper and CCW end of your reverse log pot (10k or 22k (25k nowadays) depending on op amp) to pin 8. then connect a 200ohm resistor and 330uF/6.3V BP electrolytic cap in series with the pots other terminal and connect the other side of the cap to COM.
- R3 gets changed to either 10k or 20k depending on what op amp and pot value you used.
- C6 may need to be changed as well
- upgrade C4 to a 470uF/16V BP electrolytic cap... OR just remove it
- change C1 and C2 to the biggest value that'll fit at that voltage.
Who is online
Users browsing this forum: No registered users and 67 guests