Palmer PDI-03 innards

[dai h.]

dunno if anyone is still interested but I had a look again at my Palmer innards (just the parts I can see. I don't feel like disassembling it further) :

27R0 50W alu housed wwnd x4 in parallel (so 6.75ohms DC? Measured 6.5 ohms), inductor (in series?) about 30cm dia (air core? space in middle)
"Palmer PMT14" transformer(?) with numbers on top 1,2,3...10

(all small Rs 1/4W metal film--brown band on end)
(brn-blk-blk-red 100k?) near LINE OUT jacks (x4 on back panel) with trace to 10kA LINE OUT pot (brn-blk-blk-brn 1k?) near pot, (brn-blk-blk-brn)x4 next to aforementioned 100k above, blue Siemens boxed film u47K100 next to the four Rs, 1uF100 boxed film next to that cap, (red-brn-blk-blk)R, (brn-gry-blk-brn), (org-org-blk-brn) next to BRIGHT-NORMAL-MELLOW 3-pos.switch, (brn-blk-blk-brn) next to DEEP-NORMAL-FLAT 3-pos.switch, 10kA FILTER VOLUME sw. with 47n250 boxed film, (brn-blk-blk-brn)R nearby going across chassis close to FILTER OUTS area of PCB with (brn-blk-blk-brn) R nearby, next to a Palmer PMT13 transformer? with another box that appears to be another transformer. PMT13 is the smallest (around 22x35mm), then PMT14(around 35x35mm), then the unmarked one (around 32x41mm).Back panel has GND LIFT sw., XLR and 1/4" FILTER OUTs, four LINE OUTs, SPEAKER INPUT, SPEAKER THRU, marked INPUT 8 OHMS 100W RMS. One internal heatsink about 50x200mm. Power Rs riveted to heatsink. Some white thermal transfer compound? Vents on top and bottom of unit.

[Rcubed]

Thanks Dai, very cool of you to post that info.

[dai h.]

sure no problem. Might not be very important since I think Palmer is making the 03 again. I just tried measuring the DCR with a spk. cord inserted in the spk. input (across hot and cold of the cord with nothing connected on the other end), and I'm getting about 17.4 ohms, which seems to make sense since an 8 ohm impedance is for AC, and my understanding is that somewhere around 22 ohms would be about the equivalent for DC, but maybe I measured the wrong way or something when I was inside?

oops, wait. I think I just calculated four 27 ohms in parallel without actually measuring... (crap where is that "embarassed" icon..lol)

[dai h.]

okay, I think oxidized contacts were giving me weird measurements. (After a spritzing a bit of Deoxit,) DCR across the resistors is definitely about 6 ohms, and definitely they look wired in parallel. Across the spk. input jack a bit higher around 8 ohms.

also, the power Rs are marked "5%", and "RH50" (so prob'ly 50W).

[dai h.]

what appears to be a simple R+L for the load made me think of an old article in EQ magazine, and found some info that may be of interest to anyone messing with a simple load such as that appears to be inside the PDI-03. The load in the article is made up simply of an L (two 0.5mH 2A air core inductors in parallel for a total of 0.25mH 4A) in series with an 8 ohm R (two 100W ceramic Rs for 200W peak power handling), paralleled with a simple voltage divider for a line lvl. out (235k fixed in series with a 25k level pot to jack). The "useful range" according to the article for the inductor is 0.3mH to 1.0mH (it says "per inductor", so assume paralleled and half value, i.e. 0.15 to 0.5mH?), and that "(s)maller value inductors give a tighter, more direct sound that complements ultraclean rhythms or twangy country leads", which sounded interesting.

source: "The (<$50) Speaker Emulator" by John Catheline, EQ April 1992, pp.84-85.

[dai h.]

well I've developed an interest in speaker load emulation, and have been websearching, reading the Aiken and Plate to Plate site article, various bbs archives, etc. to see if I could do anything to try to improve the sound of some of the stuff I've got (the palmer, a Marshall SE100, and Power Brake--the latter two which I really haven't used since I got the Palmer--I have used the line level input on the SE100 however). The Palmer looks like the load is part of the network described in RDH4 without the portion to emulate the low frequency resonance. It appears there is no resistor across the inductor (which I gather would be "L1" in the Aiken article) to limit the high frequency impedance rise, which if added to a Power Brake apparently helps tame some of the buzziness in the high end. There is, however, one across what would be "L1" (part "CK3" which is a 1mH inductor) on my SE100, which is a 33 ohm 25W alu housed wwnd, which if I'm interpreting the Aiken article correctly means the high frequency impedance limit is 8R2 + 33R = 41R2 (hmm... on second thought, maybe not having one across the 160uF+12.5mH raises this). The SE100 also has the low freq. resonance part of the network (with no resistor across it) which seems to make it more the more complete circuit but sounds worse (IMO) than the Palmer, which doen't seem to have that portion.

Also, I've looked into building the Aiken article spk. load emulator, and it seems to me the biggest obstacle is the 50mH inductor. It is apparently not a standard value used for speaker crossover networks therefore if you wanted a 50mH inductor, you'd have to put multiple inductors in series or get one custom made, which I think raises the price to the point where it makes more sense to buy a Weber MASS unit (or motor unit if/when they become available again) and use the motor unit as the spk. load emulator. The type of inductor (at least as described in the speaker world) also apparently affects the sound, with air core inductors generally being considered better than cored. It also looks as if you'd want a 50mH inductor to have a low DCR as well, and a 50mH air core inductor with low DCR (made of thick magnet wire) would be rather large and heavy, so the space one would take up, and the weight could be another concern. From the standpoint of easier to obtain parts, the spk. load emulator in the Plate to Plate article looks easier to reproduce since the larger inductor used is much smaller (5.6mH), and as long as a film cap wasn't used for the 470uF (going to a film cap instead of a bipolar alu electrolytic would make it expensive), cost less. Moreover, I think I could just stick the low freq. resonance portion (5.6mH + 470uF + 15R) in series with the hot on the Palmer since "Re" is just about the same (6R8 in the article vs. 6R75 in the Palmer), although the inductor value (which I assume is some smaller value meant to boost the highs) is an unknown.

some notes on the SE100 and PB:

(at least in my unit) the main load (in the Power Brake) is 10R, not 8R2, so these loads in the SE100 and PB are not exactly the same. Also, "R2" (on the Blue Gtr. PB schematic) is 100R in mine and not 56R. Re: the SE100, the loads work like this: there are 3 loads from hot to ground (see schematic heaven schematic), the first one comes in for 6dB attenuation, and the output is taken from "C". The configuration is like this:

http://www.webervst.com/lpad.htm

where R1 and R2 make up the load from hot to ground in the SE100. On 12dB reduction, the first network is switched out, and the middle one comes in. On 18dB, both come in, output taken from the middle network. On "load", the third one comes in but the two other networks can be switched in in parallel.

[12afael]

years ago I did a pcb for the main board of the se100 .

it is not tested but if some one want to build it, here it is.

http://plexilandia.googlepages.com/

Cheers
12afael

[dai h.]

hey just had a look at it. Pretty cool. Here's some pictures of mine:

http://www.geocities.com/cs_barutan/powerbrakepics.html

[RK12345]

Hello Dai,

I see you posted photos of the inside of the SE100. Can you please post or email me a photo of the VDR. Thanks

RK (qmtv@hotmail.com)