Every once in a while I see a post about Hovland and the HP-100 preamp. I worked at Hovland as a bench tech for about 3 months in 2002. I thought that since enough time has passed I would document all that I know about the inner workings of this lauded audiophile company and the esteemed HP-100 preamp.
My path to Hovland came after one month shy of 11 years employment at Alesis Studio Electronics. At Alesis I started as a Bench Tech and ended as Service Center Administrator.
During my time at Alesis I had cut my teeth on some tube preamp design. I designed the Microphone preamp that was the premier product of Requisite Audio. This mic preamp, in prototype form, was also used to record the first set of guitar amp samples that launched the Line6 Pod. Even though the world I was in was dominated by prosumer-audio recording I was a subscriber to Audio Amateur and Sound Practices Magazines to name a few, so I was aware of audiophile design techniques and philosophies. However, I was unprepared for what lay ahead at Hovland.
The interview went well. I took a little prototype of a solid state Pultec type EQ I was working on that did impress them a bit. It started out with a gent named Michael then progressed on to meeting Bob Hovland and they hired me to be the tech to do the final test and QC on the HP-100.
Now here is where things get interesting: For my training I was set to stuff the line level PCB for the HP100. The lead assembler lady explained that all of the stocked resistors have a polarity mark on them and to follow the stuffed example board and pay attention to the resistors and their polarity. Humm, I wondered. so I asked in an as pragmatic tone as I could muster "how do you determine the polarity of the resistors?" She answered "that is a secret". I just raised my eye brows in mild astonishment and smiled and followed directions. Later I commented on how I had never come across any article that ever even remotely suggested that a resister had any aspect of electronic polarity. She just kind of snapped back "Well, they do and we have to test each run of resistors because they will follow the color codes one time then might be reversed another." About a month later I saw her doing the polarity test and marking the resistors. She tuned a little credit card type portable FM radio in between stations and put on the ear buds then grabbed a resistor by one end of the leads in both hands and held it next to her chest like she was praying. Then she clasped the other lead and repeated the process. Something she was hearing on the radio was queuing her to mark one end of the resistor. I never found out any more than this about the resistor polarity marking process.
As I was stuffing the PCB I took notice that there was no apparent rhyme or reason as to the choice of resistor type used in each position of the circuit. All types of resistors were used though out the circuit including flame proof and metal oxide types. And understand, this was the main line level audio PCB not the power supply. "Wow, every resistor type under the sun" I exclaimed, while pointing to the example. The lead assembler lady said "extensive listening was done for each type of resistor in each position". I was perplexed for example that there were Caddock Bulk Foil resistors on the PBC but they were not used as plate resistors. I continued stuffing, without a schematic no less.
After two or three days I given a bench area to set up. Bob Hovland trained me on matching FET's (more on that later) and how to bring up a fresh HP-100. I was told to go to Michael to get sets of tubes. BTW Michael was my boss and he was basically the general/manufacturing manager of the place. So I go get a set of tubes from him he hands me a set, Bob walks me through the procedure. Yay! I test my first unit. I continue getting units from the assemblers, getting set of tubes from Mike, testing the preamps for about a week; Everything's fine. Then I go to Mike for a set of tubes and he says "Hold on, I ran out of tube sets". "Ok" I say.
Mike leans over toward a carton of tubes and waves his hand over them like a swami, grabs a tube and feels how heavy it is. Then he does the same with another carton again and again until he has 2 12ax7's and a 12au7. By the way these are Sovtek long plate 12ax7LPS; I can't remember what kind of 12au7. Then he fondles all three and hands them to me. While this was going on I was wondering if I was having a psychedelic flashback. I checked the walls of the room quickly and didn't notice any strange movements. This was really happening and Mike was sincere in his method of tube matching. In a state of subdued shock, I somewhat robotically take the tubes and go to my bench, install the tubes and move on with the QC.
I come to learn that this technique is referred to by Mike as dowsing. He claimed that his tube dowsing skill was verified by Ram Labs. He also dowses the Beyer Transformers for matched pairs used in the MC circuit. Also he once took a call from a company that had a batch of wire ready. He told me "I have to leave for a bit to polarize some wire". "how do you do that?" I asked. "dowsing" he answered. Mike also told me that his brother can dowse for gold. Some time later I asked him why he doesn't take his brother around the gold country and find some gold? "well, my brother has problems concentrating and can barely maintain a life in a little apartment" answers Mike. Now I am starting to feel like this is one of those calls to late night talk radio where a claim has been made that a Big Foot has been shot but the body was just too hard to retrieve from the woods.
Now I can reluctantly accept that some can be born with powers of perception that I or most don't have. But as the weeks progressed at Hovland Mike and the lead assembly lady started striking up a relationship. Mike started teaching her his dowsing technique; Then all of a sudden she was picking tube sets and the like. Now my B.S. meter is pegged fully in the red zone.
On my last day I told Mike that I did not accept that his dowsing was real. In all earnestness he understood that it is hard to believe but he steadfastly held on to his belief in his skill.
During the whole time I worked there they never gave me a schematic. Can you believe it! You are employed as a bench tech and you are never given a schematic. Being curious, I kept a PCB and a note pad off to the side of my bench and like and inmate locked away in the Tower I slowly mapped out the topology trace by trace when Mike and the Assembler Lead Lady were not watching. After about a week, low and behold! I had seen this topology before. I knew it from an article in Glass Audio Magazine called Spice And The Art Of Preamp Design. I never made note of exact values or what type resistor was used in each position. The only thing really different with the Hovland topology is they used a JFET follower between the 12au7 and the output. They claimed that they could hear differences caused by loading with just the tube follower driving the output.
The filament supply was choke filtered AC. The choke was huge. The high voltage supply was mosfet regulated. Most of the audio capacitors were, of course, Hovland Musicaps but the final output capacitor coupling the source pin of the JFET follower to the outside world was a hermetically sealed tantalum capacitor, a curious choice.
Toward the end of my time there they had the first prototype of the Voltaire isolation transformer. As good as the HP-100 sounded I had heard better and I told them this. The HP-100 had a slightly soft presentation in my opinion. Where it really fell through was instrument separation and maintaining a separate space for room reverbs to decay. Once the Voltaire power module was introduced the HP-100 came alive and Mike and Bob knew it right away. I was sitting at my bench quietly snickering saying to my self "All this chest pounding from you guys (really mostly Mike) about hundreds of hours of component selection, magic polarity assignment of resistors and wire, dowsing of tubes, twisting of wire pairs always counter clockwise (forgot to mention this before), magical layout placement of components (oh, forgot to mention this too, Mike did the PCB layout the old fashion way with tape. He claims that the placement of each part on the PCB has been listening tested also) and all the preamp really needed was very clean, balanced AC power; Gee, imagine that.
My path to Hovland came after one month shy of 11 years employment at Alesis Studio Electronics. At Alesis I started as a Bench Tech and ended as Service Center Administrator.
During my time at Alesis I had cut my teeth on some tube preamp design. I designed the Microphone preamp that was the premier product of Requisite Audio. This mic preamp, in prototype form, was also used to record the first set of guitar amp samples that launched the Line6 Pod. Even though the world I was in was dominated by prosumer-audio recording I was a subscriber to Audio Amateur and Sound Practices Magazines to name a few, so I was aware of audiophile design techniques and philosophies. However, I was unprepared for what lay ahead at Hovland.
The interview went well. I took a little prototype of a solid state Pultec type EQ I was working on that did impress them a bit. It started out with a gent named Michael then progressed on to meeting Bob Hovland and they hired me to be the tech to do the final test and QC on the HP-100.
Now here is where things get interesting: For my training I was set to stuff the line level PCB for the HP100. The lead assembler lady explained that all of the stocked resistors have a polarity mark on them and to follow the stuffed example board and pay attention to the resistors and their polarity. Humm, I wondered. so I asked in an as pragmatic tone as I could muster "how do you determine the polarity of the resistors?" She answered "that is a secret". I just raised my eye brows in mild astonishment and smiled and followed directions. Later I commented on how I had never come across any article that ever even remotely suggested that a resister had any aspect of electronic polarity. She just kind of snapped back "Well, they do and we have to test each run of resistors because they will follow the color codes one time then might be reversed another." About a month later I saw her doing the polarity test and marking the resistors. She tuned a little credit card type portable FM radio in between stations and put on the ear buds then grabbed a resistor by one end of the leads in both hands and held it next to her chest like she was praying. Then she clasped the other lead and repeated the process. Something she was hearing on the radio was queuing her to mark one end of the resistor. I never found out any more than this about the resistor polarity marking process.
As I was stuffing the PCB I took notice that there was no apparent rhyme or reason as to the choice of resistor type used in each position of the circuit. All types of resistors were used though out the circuit including flame proof and metal oxide types. And understand, this was the main line level audio PCB not the power supply. "Wow, every resistor type under the sun" I exclaimed, while pointing to the example. The lead assembler lady said "extensive listening was done for each type of resistor in each position". I was perplexed for example that there were Caddock Bulk Foil resistors on the PBC but they were not used as plate resistors. I continued stuffing, without a schematic no less.
After two or three days I given a bench area to set up. Bob Hovland trained me on matching FET's (more on that later) and how to bring up a fresh HP-100. I was told to go to Michael to get sets of tubes. BTW Michael was my boss and he was basically the general/manufacturing manager of the place. So I go get a set of tubes from him he hands me a set, Bob walks me through the procedure. Yay! I test my first unit. I continue getting units from the assemblers, getting set of tubes from Mike, testing the preamps for about a week; Everything's fine. Then I go to Mike for a set of tubes and he says "Hold on, I ran out of tube sets". "Ok" I say.
Mike leans over toward a carton of tubes and waves his hand over them like a swami, grabs a tube and feels how heavy it is. Then he does the same with another carton again and again until he has 2 12ax7's and a 12au7. By the way these are Sovtek long plate 12ax7LPS; I can't remember what kind of 12au7. Then he fondles all three and hands them to me. While this was going on I was wondering if I was having a psychedelic flashback. I checked the walls of the room quickly and didn't notice any strange movements. This was really happening and Mike was sincere in his method of tube matching. In a state of subdued shock, I somewhat robotically take the tubes and go to my bench, install the tubes and move on with the QC.
I come to learn that this technique is referred to by Mike as dowsing. He claimed that his tube dowsing skill was verified by Ram Labs. He also dowses the Beyer Transformers for matched pairs used in the MC circuit. Also he once took a call from a company that had a batch of wire ready. He told me "I have to leave for a bit to polarize some wire". "how do you do that?" I asked. "dowsing" he answered. Mike also told me that his brother can dowse for gold. Some time later I asked him why he doesn't take his brother around the gold country and find some gold? "well, my brother has problems concentrating and can barely maintain a life in a little apartment" answers Mike. Now I am starting to feel like this is one of those calls to late night talk radio where a claim has been made that a Big Foot has been shot but the body was just too hard to retrieve from the woods.
Now I can reluctantly accept that some can be born with powers of perception that I or most don't have. But as the weeks progressed at Hovland Mike and the lead assembly lady started striking up a relationship. Mike started teaching her his dowsing technique; Then all of a sudden she was picking tube sets and the like. Now my B.S. meter is pegged fully in the red zone.
On my last day I told Mike that I did not accept that his dowsing was real. In all earnestness he understood that it is hard to believe but he steadfastly held on to his belief in his skill.
During the whole time I worked there they never gave me a schematic. Can you believe it! You are employed as a bench tech and you are never given a schematic. Being curious, I kept a PCB and a note pad off to the side of my bench and like and inmate locked away in the Tower I slowly mapped out the topology trace by trace when Mike and the Assembler Lead Lady were not watching. After about a week, low and behold! I had seen this topology before. I knew it from an article in Glass Audio Magazine called Spice And The Art Of Preamp Design. I never made note of exact values or what type resistor was used in each position. The only thing really different with the Hovland topology is they used a JFET follower between the 12au7 and the output. They claimed that they could hear differences caused by loading with just the tube follower driving the output.
The filament supply was choke filtered AC. The choke was huge. The high voltage supply was mosfet regulated. Most of the audio capacitors were, of course, Hovland Musicaps but the final output capacitor coupling the source pin of the JFET follower to the outside world was a hermetically sealed tantalum capacitor, a curious choice.
Toward the end of my time there they had the first prototype of the Voltaire isolation transformer. As good as the HP-100 sounded I had heard better and I told them this. The HP-100 had a slightly soft presentation in my opinion. Where it really fell through was instrument separation and maintaining a separate space for room reverbs to decay. Once the Voltaire power module was introduced the HP-100 came alive and Mike and Bob knew it right away. I was sitting at my bench quietly snickering saying to my self "All this chest pounding from you guys (really mostly Mike) about hundreds of hours of component selection, magic polarity assignment of resistors and wire, dowsing of tubes, twisting of wire pairs always counter clockwise (forgot to mention this before), magical layout placement of components (oh, forgot to mention this too, Mike did the PCB layout the old fashion way with tape. He claims that the placement of each part on the PCB has been listening tested also) and all the preamp really needed was very clean, balanced AC power; Gee, imagine that.
Last edited:
I enjoyed your story. Just when you think you've seen it all... I joke about using a Ouija Board to calculate phase margin on one of my webpages, but then I do it the real way (math). It sounds like the people at Hovland have gone insane - gotten a bit too off balance or out of touch with reality. It's probably good that you got out before their spirits set up camp in your brain anymore than they might have.
Our local audiophile club has what appears to be a 6 foot 6 inch transgender, who sells speaker wire for $11K per 20 foot pair (if I remember correctly). "She" claims it sounds better... At a local audiophile event some egotist put on, I got an opportunity to meet the main brain behind one of the most admired no feedback hi-fi tube amps (won't mention any names or brands here); another transgender... wtf next?
Man, reading that was a trip. I always wondered if things like that went on with some of the far-out claims I've heard. I copied your post into an email for some friends that don't read this forum. Fortunately we have no girly-men in our NJAS. 
BTW, you didn't elaborate on matching fets. I assume there was no curve tracer used. More dowsing I suppose.
BTW, you didn't elaborate on matching fets. I assume there was no curve tracer used. More dowsing I suppose.You are correct - No curve tracer; But no dowsing either
. Bob Hovland had a hand wired HP-100 he called "The Mule". He set me up with the Mule matching JFETS. I just measured the voltage across a resistor. As usual they never told me what I was measuring. I have to assume that it was the source voltage of the JFET. BTW the JFET was a PN4393 if memory serves me.
So if an audio shop was organised like a bookshop the Hovland products, like some other so-called 'high end', would be in the Mind, Body and Spirit section rather than the Science and Engineering section?
We have known for years that many ebay sellers think 'matched' means 'same date code'. Now we discover that a serious manufacturer thought 'matched' means 'same tingle in my fingers'!
We have known for years that many ebay sellers think 'matched' means 'same date code'. Now we discover that a serious manufacturer thought 'matched' means 'same tingle in my fingers'!
Thanks for sharing your recollections. I had always wondered about the Hovland circuit. Judging strictly from photos of the PCB, it looked to me very, very similar to a Marantz 7 except for the 12AU7. I didn't spot the mosfet in the line stage.
I would not dismiss out of hand the "dousing" for tube selection, the wire polarity, and the resistor polarity techniques even though I haven't experienced anything like that myself. There is a lot more to audio engineering than what we can measure on a test bench.
I would not dismiss out of hand the "dousing" for tube selection, the wire polarity, and the resistor polarity techniques even though I haven't experienced anything like that myself. There is a lot more to audio engineering than what we can measure on a test bench.
Thanks, morinix, your story got me rolling on the floor.
Several years back I remotely helped someone to tune-up his HP-100 with better capacitors. Once that tantalum at the output was replaced by some film cap, the person reported that the "Hovland sound" was gone and it sounded just like any other plain vanilla preamp. Needless to say, the amp was quickly sold.
I guess they used tantalum in that location as some sort of a "harmonic synthesizer" - just look at the distortion spectrum in tantalum measurement taken by C.Bateman (below).
In the process I also reverse engineered most of the amp by looking at pictures the owner have sent me. Here's the sketch of a line stage, for those who are interested
P.S. When I saw the PCBs and all the various sorts of resistors used, my first thought was - "they must be shopping cheaply at some surplus store".
Several years back I remotely helped someone to tune-up his HP-100 with better capacitors. Once that tantalum at the output was replaced by some film cap, the person reported that the "Hovland sound" was gone and it sounded just like any other plain vanilla preamp. Needless to say, the amp was quickly sold.
I guess they used tantalum in that location as some sort of a "harmonic synthesizer" - just look at the distortion spectrum in tantalum measurement taken by C.Bateman (below).
In the process I also reverse engineered most of the amp by looking at pictures the owner have sent me. Here's the sketch of a line stage, for those who are interested
P.S. When I saw the PCBs and all the various sorts of resistors used, my first thought was - "they must be shopping cheaply at some surplus store".
Attachments
Last edited:
Is that really a line stage wth a gain of about 20, followed by a -12dB (x0.25) attenuator, then a source follower? Nobody would choose to design a line stage that way, as it increases distortion and reduces headroom.
My hunch is that the original design had less gain (more feedback) but they couldn't make it stable. Maybe 'dowsing' the prototype failed to modify the Bode stability plot, but that was the only technique the designer knew?
Remember, folks, this is expensive 'high end' stuff we are looking at. Presumably lots of glowing reviews and happy customers.
My hunch is that the original design had less gain (more feedback) but they couldn't make it stable. Maybe 'dowsing' the prototype failed to modify the Bode stability plot, but that was the only technique the designer knew?
Remember, folks, this is expensive 'high end' stuff we are looking at. Presumably lots of glowing reviews and happy customers.
Morinix said they used a PN4393. Nothing special, just a garden variety N-channel 40v TO-92. (post #6)
I suppose in order to test transformers it was required to have them clamped in a jig that was exactly rotated so the laminations lined up with the magnetic field of the Earth in "harmony". Also, the power to test these transformers had to be generated from the alternator of a 1966 Mustang that was turned by a waterwheel fed by an artesian spring on the second Tuesday of the month only.
Oh yeah, and make sure the leads are kept straight. If you bend wires it will slow down the electricity!
Its a good thing Hoveland wasn't around in the middle ages as they would have been accused of sorcery and witchcraft and burned at the stake.
Sorry, I just couldn't tsiser--i mean RESIST (wrong polarity first time).
Its a good thing Hoveland wasn't around in the middle ages as they would have been accused of sorcery and witchcraft and burned at the stake.
Sorry, I just couldn't tsiser--i mean RESIST (wrong polarity first time).
In the process I also reverse engineered most of the amp by looking at pictures the owner have sent me. Here's the sketch of a line stage, for those who are interested
You certain? Or just slandering an excellent product
It looks like a seriously misunderstood interpretation of ARC SP - 3, but at least it's easy to improve by shorting all the tubes and leaving only the fet follower. Or maybe just the tantalum cap...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.